Prevalence and associated factors of active trachoma among 1–9 years old children in Deguatemben, Tigray, Ethiopia, 2018: community cross-sectional study

The study was conducted in Deguatemben woreda, one of the 4 woredas in the Southeast zone of Tigray Regional state that has 24 kebeles (smallest administrative unit in Ethiopia) in which one of them is urban. The current population size of the woreda is estimated to be 150,254 based on the 2007 population census. The woreda has one primary hospital, 5 health centers, and 24 health posts according to the woreda data. There is no ophthalmic professional in any of the health facilities. The study was conducted from 1st March to 30th March 2018.

We conducted a community-based cross-sectional study among 1–9 years old children. All 1–9 years old children in the selected kebeles were our source of population and we include children from the age of 1–9 years old who were lived for at least 6 months in the study area. Children who are unable to undergo physical examination due to medical illness were excluded from the study.

The sample size was determined for both objectives. For the first objective, we were determined using a single population proportion formula n = (Zα/2)²p(1 − p)/d², where n is the sample size, z is the standard normal deviation, set at 1.96 (for 95% confidence interval [CI]), d is the desired degree of marginal error (taken as 0.05) and p is the estimated prevalence of active trachoma (18%) taken from similar community [15]. The minimum sample size is 227 and expecting a 10% nonresponse rate and a design effect of two the final sample size is 500. We also determined the second objective, using the associated factor of active trachoma from a study done on the previous time. Sampling was determined based on the double proportion formula on the software of Epi Info StatCalc version 7 after considering the following assumptions; 95% confidence interval (CI), 80% power, 1:1 ratio of exposed to non-exposed group, odds ratio (OR) of 2.2 and taking the proportion-1 of 55.2% and proportion-2 of 32.8% number of preschool children as a factor for active trachoma [13]. Finally, with an expected non-response rate of 10% and the design effect (2.0) the maximum sample size was 502. Then we took the largest sample size that is 502.

The study employed multi-stage with a stratified sampling technique. Initially, the stratification of kebeles into rural and urban strata was made. Seven kebeles (1 urban and 6 rural) were selected from the total kebeles in the district using a simple random sampling technique. Then, probability proportional to size (PPS) used to allocate children aged 1–9 years from each kebeles. Finally, children aged 1–9 years old were selected by a systematic random sampling method using the existing list of households’ from the family folder as a sampling frame of health extension works in the selected kebeles. In the case of more than one child 1–9 years old in the selected household, we select one using the lottery method and if in the selected household has no child aged 1–9 years old we used the next household.

Data was collected using a structured and pretested questionnaire from mothers or caregivers. Face to face interview was done with the mothers of children 1–9 years old children and direct observation was conducted to assess the availability and utilization of WASH facilities by trained data collectors.

Besides, trained ophthalmic nurses who participated in the 2013 global trachoma mapping project (GTMP) who are certified for trachoma grading examined the upper tarsal conjunctiva of each child by wearing 2.5x magnifying loupe and pen torch to assess each eye for the sign of active trachoma. Grading was done according to the WHO simplified grading system [3]. Eyelid eversion (turning out) of the children was done using an aseptic technique by using cotton tip applicators and alcohol used for hand disinfection. Active trachoma was defined as the presence of Trachomatous inflammation follicles (TF) or Trachomatous inflammation (TI) in either eye.

The questionnaire was first prepared in English and translated to the local language Tigrigna and then it was again translated back to English. Two days Training was given for both data collectors and supervisors by principal investigator and ophthalmologist. Besides, the ophthalmic nurse has received refresher training with ophthalmologists having well experienced on the grading of trachoma using the WHO grading scheme procedures and he received practical training, PowerPoint, photographic images. One week before the actual data collection period pretest was done in 5% of the sample (25 children) from a similar population outside the study area and based on the findings of the pretest, minor modifications of questions, wordings, phrases and time required to interview respondents was made. Each questionnaire collected from the field was checked for completeness, missed values, and unlikely responses and then manually cleaned up on such indications. Then data were coded and entered into a computer using Epi-Info version 7.0.9.7 for using double-entry customizing and skip benefit, then after data cleaning, it was exported to SPSS version 20 computer software packages. Data were cross-checked for consistency and accuracy.

Descriptions of the main findings were done using frequencies, percentages, and summary statistics. The binary logistic regression model was fitted to assess factors associated with active trachoma. The model’s fitness was evaluated using the Hosmer -Lemeshow goodness of fit test and if the p-value is > 0.05 that was 0.295, and then the model was considered as fit otherwise unfit. Factors with p-value < 0.25 in chi-square (X2) cross-tabulation/bivariate analysis were entered for multivariate logistic regression analysis. Multicollinearity was checked by a variance inflation factor (VIF) for VIF > 10 to be significant Multicollinearity but in our data, no variable was with VIF more than the threshold. Backward LR stepwise regression analysis was used for variables to be independent predictors for the outcome variable. Those variables with p-value< 0.05 in the multivariable analysis were considered as independent predictors for active trachoma and odds ratio with a 95% confidence interval was reported.

Clean face: a child who did not have an eye discharge or nasal discharge, fly on the face at the time of data collection [16].

Protected water source: water source protected by construction from outside contamination [18].

Active trachoma: Trachomatous inflammation follicles or Trachomatous inflammation intense [4].

Knowledge: mothers or caregivers of children 1–9 years old children were asked twelve knowledge related questions to assess their knowledge status on trachoma. Correct answers were given score 1 and incorrect answers 0. Mothers who score mean and above of the questions were labeled as “knowledgeable” and those who scored below the mean were labeled as “less knowledgeable” [19].

A total of 502 children aged 1–9 years old participated in the study making a response rate of 100%. The median age of children under study was 4 with IQR of 2–6 and more than half 266 (53%) were in the age group of 1 to 4 years old. Regarding the educational status of children, 323 (64.3%) were preschool and 172 (34.3%) were students. [Table 1].

More than one-third of 190(37.8%) of mothers had no formal education. Regarding the marital and occupational status of the caregivers, 437 (87.1%) were married and 392 (78.1%) were housewives respectively. [Table 2].

Four hundred forty-one (87.8%) mothers reported that they had heard about trachoma. Besides, more than half (53.2%) of mothers were knowledgeable about trachoma which is computed from the mean of correctly answered twelve questions asked.

Out of 502 households, the assessed latrine was available in 304(60.6%), and of those who have a latrine, 90% utilize it. More than 90% of the latrines are used by all members of the family. Only 61 and 36% of HHs had solid and liquid waste disposal system respectively. Forty-one percent (41%) of HHs had animals and only 56% had separate animal sheds. Regarding the source of energy for food cooking 314(62.5%) HHs use wood or animal dung and the rest use electric. More than 70% of HHs water source is from protected water source but more than 50% of HHs spent more than 30 min to fetch water. Only 94 (18.7%) of HHs wash their child’s face more than once per day. Three hundred seventy-two (74.1%) of mothers reported soap use for washing their child’s face but only 45.6% use always. [Table 3].

All independent variables were found statistically significant in chi-square (X2) cross-tabulation/bivariate analysis at the p-value of ≤0.25 considered for multivariate regression analysis [Table 4]. On multivariate logistic regression, the odds of having active trachoma was found five times [AOR (95% CI) = 5.01(1.79–13.96)] higher among children 1 to 4 years old than children 5 to 9 years old. The logistic analysis also revealed that the odds of having active trachoma were almost six times [AOR (95% CI) =5.84(1.79–13.96)] higher among children who don’t use soap for face washing than their counterparts. The odds of active trachoma was more than nine times [AOR (95% CI) =9.35(4.93–67.32) higher among children who do not wash their face in daily bases compared to those who wash more than once per day. Children with unclean faces were more than eighteen times [AOR (95%CI) = 18.22(4.93–69.32)] more likely to have active trachoma than children with a clean face. The odds of having active trachoma was also 0.06 [AOR (95% CI) =0.06(0.02–0.19)] times less likely among children from mothers with good knowledge than children whose mothers were less knowledgeable. [Table 4].

As the strength of our study, Tetracycline eye ointment was provided to those who had active trachoma two tubes to apply for 6 weeks. Being community-based cross-sectional and the grading of trachoma was done by certified, having more experience in trachoma grading and participated in the pre-intervention assessment of the global trachoma mapping project in the Tigray region. However, the study might have some limitations that first; the cross-sectional nature of the study design does not confirm the definitive cause and effect relationship. For those children who are older, we have not taken into consideration their practice on face washing. The questionnaires were designed and translated into a local language and re-translated to English. This could be also a recipe for bias. Furthermore, the study may prone to reporting bias since some of the data are collected based on self-reported information.