This population-based case-control study was conducted in Eastern Sennar Locality, Sennar State, central-eastern Sudan, and included 41,176 individuals. Eastern Sennar Locality is highly endemic according to surveillance databases from the Mycetoma Research Centre (MRC), a World Health Organization (WHO) Collaborating Centre on Mycetoma, and the reference centre for mycetoma management and research in Sudan.
The rationale for the is sample size is based on the assumption that the prevalence of mycetoma is five cases per 10,000 population, with the average population size of 935 people in the village, and the population of Eastern Sennar locality of 353,196 people. The calculations suggested that sampling 60 villages will give 80% power of detecting mycetoma at 5% significance level. Cases were identified through an exhaustive survey of 60/292 randomly selected villages in the five administrative units of the locality, described in more detail elsewhere [19]. The survey was conducted from June-July 2019, and every household in these sixty villages was visited. Cases were identified through clinical examination of all individuals by a medical doctor. When an individual was not at home at the examination, the household was revisited. All individuals with swelling of any part of the body or sinus formation, with or without grain discharge, were considered suspected cases of mycetoma. All suspected cases were referred to Wad Onsa Regional Mycetoma Centre, where an experienced radiologist performed lesion ultrasound examination to ascertain mycetoma diagnosis. A confirmed mycetoma case was defined as an individual with swelling or sinuses in any body part, with a pocket of fluid containing echogenic grains on ultrasound examination [20].
For each patient, three healthy controls were selected by simple random sampling. The control group was selected from households where no suspected case was detected and matched on community and sex.
After village leaders and the study population were informed about the survey objectives and process, all households in the study villages were visited. The questionnaire was written in English and was validated by a team including a medical doctor and a statistician at the MRC. Responses were captured through electronic data capture forms through open-source software called Open Data Kit (ODK), which collects, manages, and uses data in resource-constrained environments, running via Android devices. It allows for offline data collection with mobile devices in remote areas.
Household-level data included the type of material of the floor, roof, external walls of the dwelling, and hygiene and sanitation amenities. Individual-level factors included age, sex, marital status, educational level, occupation, swelling, history of trauma and wearing shoes/slippers at home and work. Further details, including clinical features, lesion onset, duration, site, and mycetoma family history, were collected from suspected cases.
Data were sent directly to a server at the MRC and imported to the Statistical Package for Social Sciences, SPSS 25 (SPSS, Chicago, IL, USA) for analysis. Descriptive analysis was performed, and bivariate analysis (chi-square test) was used to identify any statistically significant associations between explanatory variables and the outcome variable (confirmed mycetoma).
Univariate and multivariate analyses were undertaken to assess the strength of association of individual and household-level variables with disease status (i.e., being a mycetoma case or control). A stepwise forward and backward selection procedure was used to select inclusion variables in a conditional logistic regression model. P-value of 0.05 or less was used to enter variables into the model and 0.1 or above for removal from the model The strength of association of each retained variable with mycetoma was expressed using adjusted odds ratios with their 95% confidence interval (CI).The STROBE case-control reporting guidelines were used in this study [21].