Study design, period and setting characteristics
Community based cross-sectional survey was conducted in April 2019 among households with SAC (5-14 years) in seven purposively selected districts of the country. Ethiopia is administratively classified into nine regions and two city administrations. According to the 2018 demographic profile, there are 105, 350,020 populations (19); of which 51% is females, 47.3% were within the age range of 0-14 years. The total area of the country was 1,104, 300 km2 (20) and more than 80% of the population were live in rural areas (19). Looking to the weather condition of the country, Ethiopia has highland, midlan and lowland type of weather conditions.
The health policy of Ethiopia was preventive with a three tier health care delivery system (Primary health care, zonal hospital and referral or teaching hospital structure). The commonest public health problems to the populations of Ethiopia are communicable disease like malaria, tuberculosis, HIV/ AIDS, helmintheasis, chronic non-communicable disease. S.mansoni is among the commonest parasitic infections in Ethiopia.
This study was conducted in seven purposively selected districts of the country. Districts were selected based on their case report of SCH. Lists of regions with their respective districts: Errer district from Harar region, Ittang special from Gambella region, Mecha from Amhara region, Meta from Oromia which is the largest region with the highest number of population in the country, Gura ferda and Wondogenet from Southern Nations, Nationalities and People (SNNP) region and Wombera from the Benshangul Gumuz region. In each of the districts, the total population is more than 100,000 (fig: 1).
Survey procedure, sample size determination, Analysis and presentation of findings
One month prior to the MDA campaign, woreda health officers in each district were trained on how to cascade the deworming process and to train health care workers and teachers. Two weeks before administering PZQ to the SAC, two teachers from each school and one health professional from respective kebele (smaller administrative units of a district) was trained for three days by woreda health officers. Lists for enrolled SAC to be treated were taken from school register, while non-enrolled SAC were treated by doing community mobilization to come to the nearest treatment center. The number of PZQ tablets to be administered was determined using WHO dose pole. The MDA campaigns have been conducted one month prior to the present coverage validation survey.
The drugs were distributed by health care professionals and community health workers. The SAC (both enrolled and non-enrolled) were swallowed the medication in front of the drug distributor after informing all the necessary drug use indications.. One month prior to the current coverage validation survey, all the required information’s was observed, interviewed and recorded by independent monitors (IM). Therefore, the present coverage validation survey was done to validate the number of SAC who were treated in the last MDA campaign one month preceding this survey which was reported by IM. To perform this coverage validation survey, we used the sentinel site list from Federal Ministry of Health (FMOH), then seven districts purposively selected from six regions of the country: five districts undertake vertical treatment approach while two were undertaken an integrated.
The number of segments to be surveyed in each district was identified by coverage survey builder. Randomly selected segments should contain at least 16 households and the final participant households to be interviewed were selected by systematic random sampling technique after getting selection interval by dividing 50 to 16; which is K=3. Using the above sampling procedure, from seven districts a total of 3,378 households were visited and 5,679 SAC were interviewed to validate treatment coverage of PZQ against SCH in Ethiopia.
Data was collected using mobile phone data collection application of SurveyCTO. At the end of data collection, it was further transferred to STATA statistical analysis soft ware StataCorp version 14 for cleaning and analysis. Descriptive statistics was/were done and presented by table, graphs and text narration. Cross tabulation and binary logistic regression analysis was also done. Bivariate and multivariable binary logistic regression analysis was done to select potential candidate variables and to estimate the independent effect of predictors on swallowing status of PZQ and to control potential confounders’ respectivelly. Variables which satisfied the p-value criteria of <=0.25 in bivariate logistic regression analysis were taken as candidate variables for multivariable logistic regression analysis (Table: 5). Model was built using step wise backward elimination model building procedure and the effect of using model with reduced or many variables were compared by log likely hood ratio test. The instability of regression coefficient (Multicollinearty) was checked using Variance inflation factor (VIF) and the cutoff point was mean VIF greater than 10 to have significant collinearty.
The classifying ability or prediction performance of variables in the final fitted model was checked using Receiver Observed Characteristics (ROC) curve and 77.62% of PZQ swallowing was determined by SAC age, gender of SAC and school enrollment status of SAC (Fig:5). The association between dependant and independent variables was measured by AOR and statistical significance was assured by P-value < 5% (<0.05). Ethical approval to conduct the current survey was obtained from FMOH, Arba Minch University, and verbal consent was obtained from the guardian of the child.