A community-based, cross-sectional survey was conducted from August 2019 to October 2019 in two peri-urban communities and two rural communities in the Atwima Nwabiagya District of the Ashanti Region, Ghana. Household data was collected using a questionnaire and this was done immediately after the 2019 round of MDA.

Atwima Nwabiagya North District, one of the 43 districts of the Ashanti Region, was purposely selected as a study area that did not meet the expected geographical coverage of the MDA target of 80% during the 2016 and 2017 treatment campaigns (74.2% and 74.6% achieved respectively) [13]. Two rural communities in this district (Owabi and Daabaa) and two peri-urban communities (Koforidua and Ntensere) were selected for the study. The choice of rural and urban was to ensure a fair representation of the different socio-economic and cultural backgrounds. All four communities were located along the Owabi dam, and they were purposely selected for the study due to reports of high intensity of transmission of O. volvulus in these communities despite over two decades of participating in the MDA campaigns [13]

The Owabi dam is located 10 km northwest of Kumasi, the capital town of the Ashanti Region. The Owabi catchment area covers an area of 69 km2 [18] and has a population of 63,154 individuals [19]. The Owabi River flows through agricultural land close to the surrounding villages and it serves as the main source of water for farming purposes.

It is estimated that 67% of the working population in the district is engaged in agriculture. The main cash crops cultivated are cocoa, citrus, and oil palm whilst maize, cassava, plantain, cocoyam, and rice are the major food crops produced in the district. The other major economic activities are manufacturing, services, and commerce [19]. A report from the district assembly suggests high poverty levels, especially among women due to low literacy rates, and income disparities between rural and peri-urban settings [20].

The study included all persons aged 18 years and above who had lived in the study communities for more than three months before data collection. This target group was included in the study because minors (aged 17 years or below) although eligible have relatively low knowledge of the MDA exercise. Also, in Ghana, persons 18 years and above are of adult age and can accordingly give informed consent, which was a mandatory requirement for this study. The study considered persons who had lived in the community for more than three months because the most recent MDA had been conducted less than three months before this study. It was expected that the respondents would have the knowledge and the opportunity to participate in the just-ended MDA campaign, hence the decision to include them.

The awareness rate of the MDA program was projected at 41.0% [21], and a ratio of 1:1 respondent was set among being aware of the MDA program or not, and using a power of 80%, a 95% confidence interval and 10% non-response rate. A minimum sample size of 1700 was estimated. 2008 study subjects were recruited. The households in the communities (Owabi and Daabaa) were randomly selected using a simple random sampling technique. In the peri-urban communities (Koforidua and Ntensere), each community was demarcated into six clusters. Two clusters from each community were selected using a simple random technique, and then a systematic sampling technique was used to select households for the survey. Thus, in each household, one person (either the household head or a household member aged eighteen or above) was selected and interviewed.

Information collected from respondents included socio-demographic and economic factors, awareness of the MDA program, sources of information on MDA, preferred channels of information on MDA, and their perception of the benefits of the drugs dispensed with specific reference to MDA against onchocerciasis.

Data were collected using a pre-coded, structured interviewer-administered questionnaire with twenty-eight questions. An electronic database was designed using the Research Electronic Data Capture (REDCap®) system hosted at the Kwame Nkrumah University of Science and Technology [22].

Ten data collectors were trained to administer the questionnaires and conduct the interviews. Training of the data collectors included a review of the questionnaires and interview guides, field-based training, and a pilot run of the questionnaires and interview guides in an adjacent community.

The outcome variable was awareness of the MDA campaign, defined as whether the community member expressed having heard about the onchocerciasis MDA activity that was conducted in the study community during the 2019 round of MDA.

Exposure variables were the determinants of MDA awareness in a low-resource setting like Ghana. These were selected a priori based on previous studies [14,15,16]

The respondents were categorized into age groups of 18–29 years, 30–39 years, 40–49 years, 50–59 years, and 60 years and above. Marital status was categorized into single, separated, and living with a partner. Employment status was categorized into unemployed, skilled, and unskilled, and education into no formal education, basic (from primary school to junior high school), secondary (respondents who had completed senior high school), and tertiary (participants who had attained a post-secondary education).

The duration of living in the community was categorized into < 10 years and ≥ 10 years. A wealth index was constructed for the socio-economic status index of the study respondents from household asset data using principal components analysis and categorized as low and high [23].

The socio-economic variables used to determine the socio-economic status of the respondents were: electricity in the household, a bank account, radio, television, covered by health insurance, mobile phone, owning any livestock, other farm animals or poultry, refrigerator, own any agricultural land, own car/truck, sewing machine, computer, motorcycle, cement/ceramic/marble/porcelain tiles/terrazzo as flooring materials, electricity/LPG/natural gas/biogas as fuel for cooking, pipe/borehole/protected well/bottle/sachet water, flush/pour flush toilet facility, type of roofing (metal/wood/ceramic/brick tiles/slate/asbestos/sheets). These materials were coded as one if an individual has the household asset or otherwise zero.

Each asset was assigned a weight (factor score) generated through principal components analysis, [24], and the resulting asset scores were standardized to a normal distribution with a mean of zero and a standard deviation of one. Each household was assigned a score for each asset, and the scores were summed for each household; individuals were ranked according to the total score of the household in which they resided. The sample was divided into two quintiles (low and high).

Other exposure variables measured were Source of information on M DA, ever heard of onchocerciasis disease, perception of the benefits (knowing that the MDA drug prevents a disease), knowledge of onchocerciasis prevention, and participation in the 2019 MDA. The knowledge of onchocerciasis was assessed by asking sets of questions depicting the knowledge of onchocerciasis. The questions were about the mode of transmission of onchocerciasis, the body part(s) affected by onchocerciasis, signs, and symptoms of onchocerciasis, vulnerable groups most affected by onchocerciasis, and the common complication(s). These responses were scored and categorized as knowing onchocerciasis or not. Questions on the adequacy of information given to respondents were determined by finding out the extent to which the information covered when (the period) of the MDA exercise, the target disease/condition, the target population, and the benefit of the MDA exercise. The results were scored and categorized into having received enough information on the MDA (Yes) or not (No).

The data were analysed using Stata 16.0 statistical software (StataCorp. 4905 Lakeway Drive Station, Texas 77,845, USA).

Descriptive data were presented in frequency tables. Bivariate and multivariate analyses were used to measure the associations between exposure variables and the outcome variable. Results were reported as odds ratios (ORs) and 95% confidence intervals around the respective ORs. Multivariate logistic regression models were fitted using a backward stepwise approach to adjust for the effect of confounding factors.