Introduction
Onchocerciasis is a debilitating neglected tropical disease (NTD) caused by
Onchocerca volvulus, a filarial nematode transmitted by blackflies of the genus
Simulium [
1]. In 2017, it was estimated that at least 220 million people required preventive chemotherapy against onchocerciasis, 14.6 million of the infected people had skin disease and 1.2 million had vision loss [
2]. According to the Global Burden of Disease 2019 study, an estimated 19.1 million people are infected, with the disease being responsible for 1.23 [95% Uncertainty Interval = 0.77–1.82] million disability-adjusted life years (DALYs) [
3]. More than 99% of cases occur in sub-Saharan Africa [
2]. In Ghana, the at-risk population is greater than 2 million people [
4], with onchocerciasis being endemic in 15 of its 16 regions [
5].
The global health community, led by the World Health Organization (WHO) through a recently published NTD Roadmap for 2021–2030, aims at elimination (interruption) of transmission (EOT) for onchocerciasis, with 12 countries (about a third of all endemic countries) proposed to be verified for EOT by 2030 [
6]. The inspiration for this target is drawn from Sustainable Development Goal 3 (SDG 3), which aims at achieving
Good Health and Well-Being for All, the principle of leaving no one behind [
7], the London Declaration on NTDs [
8], and the recent Kigali Declaration on NTDs [
9,
10].
The achievement of onchocerciasis EOT is strongly dependent on the success of national programmes delivering annual/biannual mass drug administration (MDA) of ivermectin in endemic communities [
8]. Ivermectin is a safe and efficacious microfilaricide (i.e. clears the microfilarial progeny of the parasite), exerting also a temporary embryostatic effect (i.e. transiently reducing production of live microfilariae (mf) by the female adult worm) [
11]. Since microfilarial production is resumed within 4–6 months following treatment and skin repopulation by mf can be substantial at 12 months post-treatment [
11], annual MDA may not be sufficient to curtail transmission during the inter-treatment period, particularly in areas with high vector biting rates (e.g. hyperendemic areas), indicating the need for biannual treatment [
8]. In addition to its microfilaricidal and embryostatic effects, repeated exposure to ivermectin may lead to a more permanent sterilizing effect [
12] and/or to a macrofilaricidal effect (against the adult filariae) [
13]. However, due to the long lifespan of the latter (10 years on average), there is a need for uninterrupted high geographical and therapeutic coverage of ivermectin MDA for at least 15–20 years (and possibly longer) in order to interrupt
O. volvulus transmission in endemic areas [
14]. Successes (with annual or biannual MDA) recorded in some foci of Mali, Senegal, Nigeria, Sudan and Uganda [
15‐
18] paved the way to shifting from elimination of onchocerciasis as a public health problem (control) to EOT.
The success of ivermectin MDA programmes to achieve EOT strongly depends on sustaining high levels of treatment adherence, with systematic non-adherence (the proportion of the population never taking treatment) being one of the most important factors hindering progress [
19]. However, in MDA programmes, the frequently reported metric is the ‘therapeutic coverage’, which refers to the proportion of total (or of eligible) population who received the drug, and not necessarily ‘adherence’, which refers to the proportion of eligible population who actually ingests the drugs consistently over multiple treatment rounds [
20‐
22]. Studies in endemic communities have demonstrated that despite high reported MDA coverage, treatment adherence is far from ideal [
20,
23]. (Although the term ‘compliance’ has been used in many studies, we adopt the term ‘adherence’ to better reflect an active process of participation by individuals in treatment programmes.
1) Substantial proportions of ‘non-adherers’ can act as persistent infection reservoirs in endemic communities, hampering the achievement of EOT [
19,
22‐
24].
In Ghana, persistence of
O. volvulus infection and its associated clinical manifestations has been documented in the Bono Region despite 27 years of ivermectin treatment [
5,
25], with biannual delivery of community-directed treatment with ivermectin (CDTI) since 2009 [
26]. In fact, Ghana is a country formerly included in the Onchocerciasis Control Programme in West Africa (OCP) [
26,
27]. During the OCP (1974–2002), vector control, through aerial larviciding, was implemented in the northern and central parts of the country [
27]. When ivermectin was licensed for treatment of onchocerciasis in humans in 1987, Ghana was one of the first countries to implement MDA in 1995, via mobile teams [
5]. Following the closure of the OCP, regions classified as Special Intervention Zones (SIZ; areas where microfilarial prevalence had remained above 50%), received further interventions after 2002. In Ghana, SIZs included an area in the Pru River basin, where CDTI continued yearly till 2012 [
28].
Following decades of ivermectin MDA, an onchocerciasis impact assessment was performed in 2017 by the Ghana Health Service (GHS). Community surveys conducted during the assessment revealed that infection prevalence was unexpectedly high. In the Ottukrom and Kwanware communities in the Wenchi municipality of the Bono Region, where the microfilarial prevalence in 1995 was 54.2%, surveys revealed that the prevalence in 2017 was 29% [95% Confidence Interval, CI = 16.1–46.6%] (9/31 adults aged ≥ 20 years, examined by skin-snip microscopy) and the Ov16 seroprevalence among < 10-year-old children was 38% [95% CI = 20.8–59.1%] (8/21 children examined by Ov16 rapid diagnostic test, RDT). In Abekwae, in the Tain District, the microfilarial prevalence in 2017 was 13.9% (9/65) [95% CI = 7.5–24.3%] and the Ov16 RDT seroprevalence was 9.3% (7/75) [95% CI = 4.6–18.0%] [
5,
25]. Low treatment adherence could be a factor contributing to the persistence of infection in these and nearby communities. With the interruption of MDA in 2020 due to the COVID-19 pandemic and subsequent resumption of NTD activities in 2021, there is a need for studies to inform the implementation of remedial action, such as educational and behavioural change campaigns by the GHS, other policy-makers and implementation partners, as well as to identify mitigating strategies to help programmes get back on track to achieve the EOT target. Therefore, this study aimed to investigate the extent of and factors responsible for non-adherence to ivermectin treatment in 13 endemic communities with persistent
O. volvulus infection in the Bono Region of Ghana despite 27 years of ivermectin MDA.
Discussion
This study investigated adherence to ivermectin treatment in 13 endemic communities with persistent onchocerciasis in the Bono Region of Ghana which have been under MDA for nearly three decades.
The results showed that 36.4% of the study participants had missed at least one round of ivermectin despite being eligible at the time of distribution; 28.8% did not take ivermectin in the last (March 2021) round, and 5% had never taken the drug (systematic non-adherers). The success of onchocerciasis elimination is highly dependent on the adherence to treatment in endemic populations during MDA rounds [
8,
14,
19]. This is especially important because non-adherers, and particularly systematic non-adherers (never treated) may act as infection reservoirs in the communities, contribute to transmission and derail efforts to achieve EOT [
22]. Although the problem of non-adherence is not unusual, the observed level of non-adherence with the last round of MDA at the time of the study (29%) is relatively high. For instance, 19% of 308 study respondents reported not taking ivermectin during the last (9th ) round of annual MDA in the Kabo area (Gambella Region) of southwestern Ethiopia in 2012 [
36], and a study in Uganda reported that 21% of 839 people interviewed had not taken ivermectin during the last (10th ) round of annual MDA in 2002 in the Bushenyi District [
37]. However, the interviews in Ethiopia were conducted three weeks after MDA and those in Uganda took place two months after treatment, whilst our study was done five months after the first of the two rounds of biannual treatment in 2021, following the disruption to all treatment campaigns and NTD activities caused by the COVID-19 pandemic.
The proportion of respondents in our study who reported consistently taking ivermectin during each round since becoming eligible was 63.6%, with 79.8% taking treatment for at least five rounds. In the Bench Maji Zone of southwestern Ethiopia, the proportion of 553 respondents (aged ≥ 15 years) consistently adhering to treatment over five years of biannual MDA (10 rounds) was 65.3% [
24]. In the Centre, West and Littoral regions of Cameroon 57.8% of those interviewed (aged ≥ 10 years) declared having taken treatment each time during the last 5 MDA rounds (2010–2014), with 9.8% of systematic non-adherers (never treated) [
38].
Among the systematic non-adherers in our study, the most common reason given for never taking treatment was the fear of side-effects reported by others in the community, closely followed by being absent at the time of drug distribution. For those who did not take ivermectin during the March 2021 round, the main reason was not being in the community at the time of MDA (Table
3). Common side-effects declared during FGDs and interviews were oedema (of limbs, face, penis), boils, rashes and lesions, loss of libido, general malaise, musculoskeletal pains, immobility, dizziness and headaches. Most participants, however, admitted that these side-effects often resolved within 2–5 days. Fear of side-effects has also been reported as a major reason for not adhering to treatment in previous studies conducted in Ghana [
23,
39]. In the Upper Denkyira East Municipality, Central Region, fear of side-effects was reported by a decreasing fraction of respondents (76%, 23%, 16%) in 2002, 2006 and 2013, respectively, but the proportion taking treatment did not follow a correspondingly increasing trend (76%, 64%, 79%), yielding a mean rate of non-adherence to ivermectin intake of 27% for those years [
39], in line with our 29% for 2021.
Experiencing severe adverse effects (SAEs) in past treatment rounds, or fear of SAEs experienced by others negatively impacts on community participation in and treatment adherence to ivermectin MDA, particularly in onchocerciasis-loiasis co-endemic areas [
19,
21,
40,
41]. A study in South-West Cameroon demonstrated that the fear or past experience of side-effects associated with ivermectin treatment was the main reason for non-adherence despite the fact that the area was at relatively low risk of loiasis and that no fatal encephalopathies had been reported [
19]. Wanji et al. [
40], also working in onchocerciasis-loiasis co-endemic areas of South-West Cameroon, documented that the proportion of systematic non-adherers was nearly 16%, and that although the majority (40%) of the study participants (2,364 people) had taken the drug 1–3 times, only 18% had taken it at least 7 times (quantified by the participants’ oral declaration). There was also a clear correlation between treatment adherence and levels of microfilarial infection, with the highest prevalence (60%) found among the systematic non-adherers and the lowest (34%) among those who had taken ivermectin ≥ 7 times [
40]. A relationship between treatment adherence and microfilarial prevalence was also reported in the Bench Maji Zone of southwestern Ethiopia (without
Loa loa), where the prevalence among those who had missed at least one MDA round for the past 15 years of CDTI was 10% compared to 3% in those who had consistently taken ivermectin. The proportion of participants refusing treatment (systematic non-adherers) was 5.6% (31/553) [
42].
Absenteeism due to travelling was also one of the most frequently recorded reason for not taking ivermectin in our study area (both among systematic non-adherers and among those who had missed the March 2021 treatment round, Table
3). In fact, most community residents had not been not born in the villages (54.2%, Table
1) but had made their settlements in the communities for various reasons, such as farming. Such persons usually return to their ‘hometowns’ during festive periods and also after major farming seasons. In a study by Hamilton et al., in the then Brong Ahafo region of Ghana, absence during drug distribution was the major contributor (52%) to missed treatments [
43]. Being absent at the time of MDA was also reported in Cameroon and two studies in Ghana, where it was given as a reason for not taking ivermectin in, respectively, 37%, 32% and 30% of those who had missed the last treatment round being evaluated [
21,
23,
27]. The study by Senyonjo et al. [
21] in Cameroon remarked that the lowest levels of adherence, recorded for young adults (aged 20–34 years), could be due to increased work and mobility amongst this age-group, while lower adherence levels among those who had moved into the village in the last five years compared to longer-term residents, could be owing to lack of awareness of the MDA campaign and/or the risks of onchocerciasis.
In our study, age was not significantly associated with treatment adherence (Tables
4 and
5), in contrast with other studies which found age to be positively [
38] or negatively [
36] associated with taking ivermectin. Our study found that men were more likely to have taken ivermectin compared to women (Table
4), in agreement with Agyemang et al. [
39], also in Ghana, whilst no statistically significant differences in adherence between males and females were reported in other studies [
19,
20,
23,
24,
38,
43]. Our lower adherence among women contrasts with the notion that women tend to have better health-seeking behaviours than men [
44]. However, for young women of reproductive age (WRA), a major reason not to take ivermectin was pregnancy. Some women expressed concern that because they frequently undergo the cycle of pregnancy, motherhood, breast-feeding and back to pregnancy, they missed several ivermectin treatment rounds. In the study by Forrer et al., refusal of ivermectin by WRA based on the belief that it leads to miscarriages was highlighted as a factor contributing to non-adherence in this population group [
19]. To date, pregnant women are excluded from MDA with ivermectin programmes for onchocerciasis and other helminthiases (such as lymphatic filariasis). Ivermectin has been assigned to pregnancy risk category C (risk cannot be ruled out) [
45] by the USA Food and Drug Administration (FDA), and therefore the manufacturers consider ivermectin contraindicated in pregnancy. However, in their systematic review of the safety of oral ivermectin during pregnancy, Nicolas et al. found that no study reported neonatal deaths, maternal morbidity, preterm births or low birthweight [
46]. Exclusion of pregnant women may help sustain a substantial infection reservoir and deprive a vulnerable population of potential benefits, as there are indications that treating
O. volvulus-infected women may improve pregnancy outcomes and reduce the risk that their children develop onchocerciasis-associated morbidities [
47]. Therefore, further studies are needed to investigate the safety and potential benefits of ivermectin for this vulnerable population. It should also be investigated whether, when taking into account pregnancy, being female remains significantly associated with lower ivermectin adherence.
Another reason for non-adherence identified in our study was the desire to drink alcohol, especially among men. Alcohol consumption in some of the study communities was relatively high as it is usually considered
‘men’s water’ in some villages. A typical routine for some young men in several communities is to ‘
cut a little alcohol’ for appetite, libido and energy to work. In the Democratic Republic of Congo (with co-endemic loiasis), alcohol intake 24 h prior to ivermectin treatment was significantly associated with neurological SAEs [
41]. However, in a study by Homeida et al., a locally brewed alcoholic beverage given with ivermectin did not cause changes in the plasma pharmacokinetic parameters of ivermectin, suggesting that alcohol intake is unlikely to be a contributory factor in the development of the SAEs that can occur following ivermectin treatment of individuals co-infected with loiasis [
48].
Our multivariable analysis indicated that individuals with better perception and knowledge of the disease and the beneficial effects of MDA had better health-seeking behaviours and were more adherent to ivermectin intake. These results were echoed in both the quantitative and qualitative findings of our study, and are in agreement with the results of other studies [
23,
38]. We also found that people with lower levels of formal education tended to be more adherent to MDA than those who had attained higher levels (Tables
4 and
5). This finding is consistent with a study conducted in Cameroon and Nigeria, which discussed that those with higher education may be more mobile and harder to reach during MDA campaigns [
49]. For instance, people who are undergoing formal education in secondary and tertiary schools (mostly located in urban areas) may have to travel outside their communities for several months of the year for school, increasing their chances of missing MDA rounds.
The low CDD : population ratio (about 1 : 1,000) in our study communities is of concern (Table S2). The travel time between communities explains why certain individuals may not be reached. The reasons for the low numbers of CDDs in endemic communities in Ghana may include lack of cooperation and/or multiple demands by community members, scarce resources for the work and insufficient financial incentives, leading to the resignation of CDDs [
32]. The study of Agyemang et al. [
39] in Ghana pointed out that CDDs are required to complete the distribution of ivermectin in the entire village, covering all households, within just seven days irrespective of the size of the catchment area. As the success of MDA depends, to a large extent, on the essential role of CDDs [
32], it is important to put in place remedial actions to address these challenges.
This study had a number of limitations. Study participants were not randomly selected. Therefore, results need to be interpreted with caution as the findings may not be representative of all the persons living in the study communities. Also, a social desirability bias may have influenced the responses to the questions, although participants were made aware that the responses had no punitive implications for them whatsoever. To reduce recall bias, we performed the study five months after the last round and immediately before the next round (which took place in August 2021 in an effort to mitigate the impact of the missed MDA treatment rounds due to COVID-19). However, other adherence studies were conducted much sooner after the last treatment round being evaluated [
36,
37]. Notwithstanding, the particular features of ivermectin tablets (small-size, white-colour tablets), their availability only during MDA campaigns, and their distribution accompanied by sensitization and door-to-door delivery, provide a strong reference for people to recall [
50]. Another limitation is that we did not interview healthcare workers at district, municipal, regional and national levels, as we focused on community-level factors that affect treatment adherence. Finally, we did not interview community-based organizations, which might have provided valuable information on treatment adherence in their communities.