Ethics statement

Ethical clearances for this study were obtained from the Ghana Health Service Ethics Review Committee (Clearance No GHS-ERC: 006/09/23). Additionally, administrative authorization was granted by the Director General of Ghana Health Services. Community consultation, mobilization, and raising awareness were conducted, during which verbal consent was obtained from community leaders. All participants underwent an informed consent process, and written informed consent forms were signed before data collection. For individuals younger than 18 years, assent forms were signed by their parents or guardians. The study adhered to ethical standards in handling participant data and ensuring voluntary participation.

Settings

The 10 km radius area of the Kwanware-Ottou transmission focus covers parts of Subsinso, Nsawkaw, and Boase sub-districts in the Wenchi, Tain, and Banda HDs of Ghana, respectively. This area is drained by a dense river network, with two main rivers, Subin and Tain, both having actual and potential blackfly breeding sites [10,11]. The main occupations of people living in these districts are cash crop farming, particularly cashew nuts and fruits. Additionally, artisanal mining is a major activity in the Nsawkaw sub-district, while cattle rearing is primarily practiced in the Subinso and Boase areas in Wenchi and Banda HDs, respectively. The intervention was implemented at the sub-district level to align with health system smallest operational unit. These sub-districts have well-defined geographic boundaries for programme implementation and include all communities within or near the 10 km radius around Ottou community, ensuring comprehensive coverage of the focus. See Fig 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Map of the study area showing a 10 km radius onchocerciasis transmission focus area around Ottou community.

Parasitology findings, shown as pie charts, indicate the proportion of individuals positive for microfilariae (mf+) detected via skin snip (March 2021). Serology findings are represented by colour-graded squares, with size and shade reflecting Ov16 antibody prevalence (March 2021). Star symbols mark blackfly collection sites, with dark blue stars indicating detection of infective L3 larvae (October 2020). Coloured circles show locations where blackfly larvae were found. The size is proportional to the number of larvae detected. This map has been modified from our previous publication [11]. The base map uses national and district administrative boundaries from the geoBoundaries Global Database (https://www.geoboundaries.org/countryDownloads.html; license: CC BY 4.0, https://www.geoboundaries.org/metadata.html). Additional data sources include HydroATLAS hydrography (https://www.hydrosheds.org/hydroatlas; CC BY 4.0), manually digitized roads from high-resolution satellite imagery, and primary field-collected population and survey data. The map was created in ArcGIS Pro software using only openly licensed datasets.

https://doi.org/10.1371/journal.pntd.0013958.g001

Study design

A mixed-method approach (quantitative and qualitative) was conducted from January to November 2024 to explore the implementation of CDTI in the study area, identifying challenges and solutions. This approach combined community consultations, Geographical Information System (GIS) technology, census including treatment coverage evaluation surveys, and qualitative assessments to elucidate context-specific challenges and solutions. These findings were subsequently used to guide the participatory development of an Action Plan (AP) aimed at improving CDTI effectiveness. The AP was then implemented by the programme. Fig 2 outlines the study design and procedure and Table 1 shows activity timeline.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Overview of activities and timelines.

https://doi.org/10.1371/journal.pntd.0013958.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Study design and procedure.

All steps involved strong community engagement and participation.

https://doi.org/10.1371/journal.pntd.0013958.g002

Data management and analysis

A GIS expert verified completeness and accuracy of GIS locations as they were being collected. Following field collection, data were downloaded and cleaned with confirmed settlements mapped. Census data was downloaded from the Commcare platform and cleaned and analysed using STATA statistical software (StataCorp. 2023. Stata Statistical Software: Release 18. College Station, TX: StataCorp LLC). The population of each settlement was estimated from the census. Logistic regression was used to estimate the association (odds ratio, OR) of each variable with the outcome of interest (treatment) at 95% confidence interval (CI). P < 0.05 was considered statistically significant.

During qualitative data collection, data collectors held daily meetings to discuss and synthesize findings, identify areas for further exploration, and determine points of saturation. Interviews were recorded using digital Dictaphones and later transcribed verbatim. To ensure quality of transcription, research assistants that conducted the interviews also performed the transcriptions [19]. However, all transcripts were independently reviewed and verified by a second researcher to ensure accuracy and reduce interpretive bias. Analysis followed a thematic approach, developed based on participant responses and study objectives, incorporating both deductive and inductive elements to capture emerging insights. NVivo software (version 14) was used to organize and manage the coding process. Verbatim quotes were selected to best illustrate key themes aligned with study objectives. This was used to triangulate quantitative findings from both programme-reported and survey-derived coverage data. Post-implementation programme coverage was compared to pre-intervention survey coverage to assess changes attributable to the intervention package.

1. Identification of settlements through community consultations

During community consultations, 10 hamlets – small remote settlements (Ampiani, Kwaedenden, Bofuorkurom, Kwadom, Kwadom-Nsuta, Kwadom-Boase, Winamda, Branam State Farm, Yibong-Akura, and Mambeele) were identified and included. These had all been included into larger communities during preceding CDTI round.

2. Identification of settlements from satellite imagery

The outcomes of the first few days of field verification were useful for data collectors to prioritize potential settlements to visit subsequently. Of 479 locations identified on satellite image, 265 were visited from March 16–24, 2024. Local guides confirmed that remaining unvisited locations were abandoned or temporary shelters. Households were mapped as unique settlements based on their clustering. Combining GIS and census data, 19 named settlements with inhabitants were identified. Fig 3 maps these settlements within the 10 km transmission focus. Of these settlements, 12 did not have own community treatment registers during preceding CDTI which masked the need for CDDs as well as straining their ability to reach them.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Map of verified settlements.

The circle indicates a 10 km radius transmission focus, defined based on parasitological, serological, and entomological data. These data show a peak in transmission intensity at the center of the circle, gradually declining to undetectable levels at the 10 km boundary [11]. Data Sources and References: The base map uses national and district administrative boundaries from the geoBoundaries Global Database (https://www.geoboundaries.org/countryDownloads.html; license: CC BY 4.0, https://www.geoboundaries.org/metadata.html). Population density and survey results were obtained from primary field-collected data. Additional data sources include HydroATLAS hydrography (https://www.hydrosheds.org/hydroatlas; CC BY 4.0), manually digitized roads derived from high-resolution satellite imagery, and other primary field-collected population and survey data. The map was created in ArcGIS Pro software using only openly licensed datasets.

https://doi.org/10.1371/journal.pntd.0013958.g003

3. Census and treatment coverage evaluation

A total of 3,338 individuals irrespective of their treatment eligibility were censused in the 19 communities/settlements. In March 24, 70.8% (Confidence Interval, CI: 69.2-72.4) of the total population indicated they swallowed ivermectin in the preceding CDTI carried out 1–2 months earlier. Programme reported a coverage of 88.9% which falls well above the upper limit of the survey coverage’s 95% CI. Eight communities, however, reported coverage levels of less than 65%, with the lowest reported in Papaase (25.8%, CI: 13.5-43.8), Kwanware (46.4%, CI: 36.7-56.3), Mensakrum (59.6%, CI: 53.6-65.2), and Gyabaa (46.9%, CI: 33.5-60.8).

A total of 548 eligible participants (20%) were not offered ivermectin. When grouped into main (large and accessible) and satellite (remote with poor accessibility) communities, the difference in the proportion of people not offered ivermectin was significant (15% vs. 24%, respectively; p < 0.001). All 466 individuals (14.0% of total population) who reported not registered during the preceding CDTI also indicated that they were neither offered nor took ivermectin, as expected. Thus 85.0% (466/548) of those not offered ivermectin were not registered. Among those not offered ivermectin, 30% (134 out of 452 respondents) reported that CDDs did not visit their communities. Eight out of ten individuals who were offered but did not swallow ivermectin cited fear of adverse effects as the reason. Among those who took ivermectin, 3.1% (73 individuals) reported experiencing adverse effects with the most frequent being itches. See Fig 4.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Complaints after ivermectin intake among 73 respondents.

https://doi.org/10.1371/journal.pntd.0013958.g004

Other reasons for not taking ivermectin included absenteeism and communication issues. Among the 47 individuals who did not take the medicine due to being absent, 22 (46.8%) were at work or school, and 16 (34.0%) were traveling outside the community. Additionally, 167 age-eligible individuals reported not hearing about the MDA. A slightly higher proportion of individuals in satellite communities 96 (6.6%) were unaware of the MDA compared to those in main communities 71 (5.5%; p = 0.23). Regarding satisfaction, 65 participants expressed dissatisfaction: 54 had not received ivermectin, and 7 cited adverse effects. Among those dissatisfied, 75% (49 out of 65) suggested that CDDs should visit them or their communities to offer the medication.

In a multiple logistic regression, compared to those whose main occupation was farming, individuals whose main occupation was mining were less likely to have swallowed ivermectin (OR: 0.26; 95% CI: 0.10–0.72). Regarding movement histories, those visiting the community (0.18; CI: 0.08 - 0.40) and permanent residents who had travelled outside the district (0.26; 0.17 – 0.38) were both less likely to have taken IVM compared to permanent residents who had not travelled outside the district in the past twelve months. Similarly, shorter stay in the community was associated with lower OR of taking ivermectin. See Table 2. May, June, and July were reported as months of least movement out of the community.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Ivermectin Coverage Evaluation by Settlement.

https://doi.org/10.1371/journal.pntd.0013958.t002

4. Qualitative assessments identified specific challenges and solutions

Key findings from the qualitative assessment revealed context-specific challenges and solutions aimed at improving participation in CDTI. These issues were then considered during development of the AP. Key thematic issues are highlighted below.

Misconceptions about onchocerciasis and its mode of transmission

Awareness refers to the accurate information that community members have about onchocerciasis and CDTI process—its purpose and benefits—which, influence their engagement and participation [20]. Awareness is influence by misconception. Misconceptions about onchocerciasis transmission among some community members included the beliefs that the disease is gotten from a contaminated water source or mosquito bite.

“The disease is from the river we drink from the streams; people urinate into the water while animals drink the water too. We need to tackle that first else we can take the drug for thousand years and still get the infection”. (FGD with adult men, Wenamda)

As solutions, participants from different communities recommended increasing awareness about the importance of the drug, its possible adverse effects, and how to manage these adverse effects through community meetings. “I know we have not had such a gathering since I came to this community 22 years ago.” The participants believe it will encourage community participation in the CDTI going forward:

“I would suggest that we hold a community meeting or ‘durbar’ where the experts and elders’ team up with the agent to educate us. A general meeting would help bring out necessary questions or information to learn from which will help us all.” (KII with parent of never treated individual, Branam)

Participants suggested that the community should be informed of the treatment in advance to ensure people are available during the CDTI.

“They do not inform that they will be distributing the medicine ahead of time. Inform us ahead of time! We can organize the community for the effective distribution of the medicine.” (FGD with men, Kwadom Joe-boy Community).

“You could sometimes give the drug to someone, and they refuse to take them because they do not know its use.” (KII with CDD, Kwadom Joe-boy Community)

High migration of community members

One challenge for implementing CDTI in these communities is the high level of migration during the farming season. CDDs reported that when distribution is during rainy season, they are unable to access some community members who might have moved temporarily to farms. In addition, they stated that some people’s nature of work demands that they travel away from the community for a certain period.

“I am usually absent anytime they come to distribute the drugs. And when they come, they give to available persons only. They do not give to you other people’s drugs to keep for them.” (KII with never treated person, Branam)

“There are some places where the person (referring to the CDD) will tell you that your name is not in their records hence they cannot give you. For me, I haven’t taken the medicine for a while, and it is because I travel.” (FGD with adult men, Kwadom Nsuta)

Low participation of cattle herders

Cattle herders—particularly semi-nomadic individuals from the Fulani ethnic group—reported that they were often absent during MDA campaigns due to their daily mobility and seasonal migration patterns. However, we recognize that not all individuals engaged in cattle rearing are nomadic. The cattle headers occupation category in Table 3 had relatively high coverage (71.7%). This category likely includes settled cattle herders who are more consistently present in the community and therefore more likely to participate in CDTI.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Factors associated with participation in Community-Directed Treatment with Ivermectin among participants who were 15 years or older.

https://doi.org/10.1371/journal.pntd.0013958.t003

Lack of resident community drug distributors

Some communities lack resident CDDs.Hence, CDDs from other communities visited to distribute ivermectin. Community members reported that the absence of resident CDDs led to many community members missing treatment as they were absent when the CDD visited their communities. They requested for CDDs to be selected from their communities to ensure everyone is treated during CDTI:

“The distributor comes to the community during the day when everyone has gone to farm, so many did not get the drug but if the distributor is resident in the community, he will distribute the drug in the evening when everyone will be back from the farm and then no one will be left out” (FGD with Men, Kwadom Nsuta)

Community members also suggested that the number of CDDs should be increased.

“So, if you could add more CDDs, it could help to do effective work. When you look at the Boase village and this place and the other surrounding villages, it shouldn’t be one person alone sharing for all these villages.” (FGD with adult men, Kwadom Nsuta)

Lack of training for some Community Drug Distributors

CDDs identified a lack of training as an issue affecting effective delivery of medicine to the community. For instance, one CDD said he “had received no formal training although he had been distributing drugs for five years”. He “relied on personal discretion to select who was offered the drug and the dosage”. This concern was also identified by a community head in another community, reporting that CDDs “no longer measured the height of individuals before distributing ivermectin”, because of an absence of training. He raised concern that dosages could be incorrect.

“Initially they used to measure us and so from there, he knows that I am X inch and so he must give me four [tablets]. But now, he uses his eyes to do it.” (KII with Community Head, Kwanware)

Directly observation therapy (DOT) not practised

Another challenge of implementing CDTI identified in some communities is that DOT, recommended by the programme, is not strictly adhere to. Community members reported that CDDs distributed the drugs without ensuring it was swallowed, yet noting the individual as treated in the register:

“Initially when they started the distribution, they ensured that people took the drug in front of them [CDDs], but now they have stopped. So, it makes it difficult to know those who have actually swallowed the drug.” (FGD with men, Wenamda).

“Some people collected the drug and did not swallow it but kept it in their rooms and you won’t know.” (FGD with adult men, Kwadom Nsuta)

Community members therefore recommended that DOT should be practiced. In doing this, persons who do not take or refuse to swallow the drug would be easily identified for action.

Refusal to take the medicine due to adverse effect

A major reason given by participants for low participation in CDTI in the communities is the fear of adverse effects after actual experience or observation in others who swallowed the medicine. Women, mostly from Kwadom Nsuta settlements, narrated their ordeal after taking the medicine and decided they would not take again.

“About three years ago, I was given oncho [onchocerciasis] drug [ivermectin], and my body got swollen. I when to the hospital several times. I was given injection, but the condition was the same. The condition is no more with me but, if they bring the oncho drug again, I will not take.” (FGD with women, Kwadom Nsuta)

“No…. when they brought the drug, I too, did not take. As for oncho truly speaking, I nearly died, so, I will not take again.” (FGD with women, Kwadom Nsuta.

5. Participatory development and implementation of Action Plan (AP)

The AP development workshop was highly participatory and interactive, engaging a broad range of stakeholders including health administrators, District NTD Focal Persons, Sub-district representatives, and community leaders and members. All participants attended and actively contributed. CDDs appreciated the involvement of opinion leaders and local health authorities in jointly identifying solutions. They noted that this collaborative engagement was the first time and that the approach mobilized effectively community and changed community’s perceptions of CDTI from being solely a CDD responsibility to that of the whole community. Community members valued the opportunity to have their voices heard and considered in shaping CDTI activities.

The AP specified activities and the responsible parties for each activity, outlined reasonable timelines, and identified the resources and capacities needed for implementation. Key components of the AP included strengthening local advocacy and resource mobilisation, engaging with hard-to-reach groups including semi-nomadic cattle herders and artisanal miners, conducting community outreach to raise awareness, shifting CDTI period to July (peak immigration), and recruiting and (re)training all CDDs on awareness raising, census and (re) registration of community members using new paper register and electronic platform as well as ivermectin administration using dosing poles. Although community members expressed the need for additional CDDs, the final number was determined by the programme during AP development, based on feasibility and operational needs. This decision considered factors such as population size, settlement patterns, accessibility and resources. During the implementation of the AP, field experiences and adjustments were recorded in an activity log. Summary of the AP implementation process is found in S3 Text and the district refined AP are found in S4-S6 Tables.

6. Evaluation of action plan implementation

This was achieved through both qualitative and quantitative assessments.

6.1. Qualitative assessment of action plan implementation

Participants indicated that significantly more people were treated compared to the previous CDTI round. They appreciated the various actions taken that encouraged greater community participation.

“Yes, they came to inform us that they would be bringing the drugs. This time, there was someone in the community [a newly recruited CDD], and he made sure everyone who was eligible to take medicine got it.”

“Yes, there was enough medicine for everyone. I did not hear of any medicine shortage.” (FGD with men, Kwadom Nsuta)

Furthermore, participants appreciated the timing of the MDA, stating that it corresponded to when people were available.

“I also like the time. Blackflies bring the oncho disease, so during the rainy seasons, that is when the flies are dominant and people are around, so this was the same time they brought the oncho drugs.” (FGD with women, Kwadom Boase).

Lastly, the acceptability of the programme had improved among the community members based on their new understanding of the importance and benefits of the medicine.

“Initially, community members were a bit hesitant to take the medication, but this time, when it came, people came for it in a rush. This time, almost everyone, both children and adults, was aware of the medication.” (KII CDD, Ottou)

Participants also highlighted areas for further improvement. Notably, the request for a follow-up to treat those who were missed.

“When they left, some people came saying that they did not get the drugs. So, I am pleading that as you go, tell them not to delay coming and share for those who did not get it. When they came, their wives and children were not around. So, they should come back and give them the drug”. ((FGD Kwadom Boase men)

6.2. Quantitative assessment of action plan implementation

The sub-districts have well-defined geographical boundaries and serve as the primary level for CDTI programme data collation. Therefore, we consider this suitable geographic scale for evaluating the impact of the intervention. For the three sub-districts included in the intervention, the old registers recorded 6,898 more individuals. This is 13.1% higher than the new register record (59,710 vs. 52,812) even though the new registration was during the peak immigration (rainy) season. Since the intervention was closely monitored, the treatment coverage data obtained was reliable and comparable to the survey coverage recorded during the initial census. For the population in mining settlements which were located entirely in the focal area, a smaller population was counted during census compared to the new registration - 474 vs 370— 21.9% less. Conversely, farming settlements saw an increase from 707 to 886, 25.3% higher. Coverage increased by 17.4% (70.8% vs 88.2%). All communities/settlements achieved and exceeded the 65% recommended coverage, indicating a substantial improvement across the board. See Table 2 above.

7. Refinements of the action plan

Field experiences documented in the activity log, along with qualitative assessments and quantitative analysis of reported treatment coverage across the three sub-districts, informed the refinement of the AP. The refinement was carried out during a workshop involving programme implementers and community stakeholders. They reflected on successes, challenges, and opportunities for improvement. Key challenges identified included limited interest among youth in serving as CDDs, short CDTI periods, delays in fund transfers from national to district levels, high mobility of miners and nomadic cattle herders, and language barriers—particularly with French-speaking artisanal miners and some herders. These challenges and solutions were reflected in the AP. Despite the operational difficulties, programme implementers commended the AP development and implementation as a viable and adaptive approach for continuously addressing field-level issues and enhancing programme delivery. Further details are provided in S3 Text & S4-S6 Tables.