Socio-demographic characteristics of study participants
Among 5,979 SAC (5-14 years) eligible for interview in seven districts, 5,679 SAC were interviewed, yielding a response rate of 95%. The mean age of SAC was 9.6 years with standard deviation of ±2.8. Almost male and female were equally involved in the survey. Concerning to the school attendance, 4750 (83.6%) of children attended school; of which 4732 (99.6%) children attended at primary school owned by public (Table 1).
Table 1: Socio-demographic characteristics of SAC in study districts, Ethiopia, 2019
S. No
|
Variables
|
Categories
|
Frequency
|
Percent (%)
|
1
|
Age (Years)
|
5-9
|
2765
|
48.7
|
10-14
|
2914
|
51.3
|
2
|
Gender
|
Female
|
2853
|
50.2
|
Male
|
2826
|
49.8
|
3
|
School Attendance
|
No
|
929
|
16.4
|
Yes
|
4750
|
83.6
|
4
|
Educational level
|
Primary
|
4732
|
99.6
|
Secondary
|
18
|
0.4
|
5
|
School type
|
Public
|
4612
|
97.1
|
|
|
Private
|
126
|
2.6
|
|
|
Religious
|
12
|
0.3
|
Praziquantel treatment coverage against SCH among SAC in study districts, Ethiopia
Out of 5,679 interviewed SAC in districts undertake vertical and integrated mass drug treatment approach, the overall treatment coverage of Praziquantal against SCH was 4286 (75.5%). From 4571 interviewed SAC in districts undertake vertical treatment approach, 3316 (72.5%) of SAC were reported that they were treated with PZQ; and among 1108 SAC interviewed in districts undertake integrated treatment approach, 970 (87.5%) were treated by PZQ against SCH. District disaggregated treatment coverage of Praziquantal showed that students from Itang special in Gambella region reported the highest treatment coverage (90%) and SAC from Errer district in Hareri region had the lowest (59.7%) treatment coverage (Fig:2 ).
Treatment coverage of PZQ disaggregated by gender, age, school attendance and heard MDA in the sampled districts, Ethiopia, 2019
Gender disaggregated treatment coverage of PZQ showed that male SAC were significantly more likely to consume the drug (77.6%) than females (73.3%) (X2 =14.77, p-value<0.001); treatment coverage of SCH was significantly different between age groups of SAC. SAC with 10-14 years swallowed the drug better (83.7%) than 5-9 years (66.8%) (X2 =222.66, p-value <0.001). Statistically significant difference was observed on swallowing status of PZQ among SAC who attended school and heard about MDA when compared with their counter parts (Table: 2).
Table 2: Treatment coverage of PZQ among SAC disaggregated by gender and other predictors in the sampled districts of Ethiopia, 2019
Variable
|
Categories
|
Swallowed PZQ
|
Chi-square
(p-value)
|
Yes
|
No
|
Unknown
|
No
|
%
|
No
|
%
|
No
|
%
|
Gender
|
Female
|
2092
|
73.3
|
679
|
23.8
|
82
|
2.9
|
14.77 (0.001)
|
Male
|
2194
|
77.6
|
556
|
19.7
|
76
|
2.7
|
Age (Years)
|
5-9
|
1848
|
66.8
|
799
|
28.9
|
118
|
4.3
|
222.66 (0.001)
|
10-14
|
2438
|
83.7
|
436
|
15
|
40
|
1.4
|
School attendance
|
Yes
|
4091
|
86.1
|
616
|
13
|
43
|
0.9
|
1833.40 (0.001)
|
No
|
195
|
21
|
619
|
66.6
|
115
|
12.4
|
Heard about MDA
|
Yes
|
413
|
91.3
|
371
|
8.4
|
9
|
0.2
|
2696.39 (0.001)
|
No
|
271
|
21.1
|
864
|
67.3
|
149
|
11.6
|
SCH treatment coverage disaggregated by school attendance in districts undertook vertical treatment approach
The school attendance disaggregated analysis in districts conducted vertical treatment approach showed that PZQ treatment coverage among SAC was significantly higher for SAC who attended school (84.1%) compared to their counter parts (14.4%) (X2 =1595.32,p-value<0.001) (Table: 3).
Table 3: Treatment coverage of PZQ disaggregated by school attendance and other predictors for districts applied vertical treatment approach, Ethiopia, 2019
Variable
|
Categories
|
Swallowed PZQ
|
Chi-square
(p-value)
|
Yes
|
No
|
Unknown
|
No
|
%
|
No
|
%
|
No
|
%
|
Gender
|
Female
|
1624
|
70.2
|
614
|
26.6
|
75
|
3.2
|
12.97 (0.002)
|
Male
|
1692
|
75.0
|
500
|
22.1
|
66
|
2.9
|
Age (Years)
|
5-9
|
1405
|
63.4
|
709
|
32
|
103
|
4.6
|
186.19 (0.001)
|
10-14
|
1911
|
81.2
|
405
|
17.2
|
38
|
1.6
|
School attendance
|
Yes
|
3207
|
84.1
|
566
|
14.9
|
39
|
1.0
|
1595.32(0.001)
|
No
|
109
|
14.4
|
548
|
72.2
|
102
|
13.4
|
Treatment coverage of PZQ among districts involved vertical and integrated treatment approach
Swallowing status of PZQ against SCH significantly higher among SAC in districts conducted integrated treatment approach compared with vertical (Table: 4).
Table 4 : Treatment coverage of PZQ disaggregated by treatment approach, Ethiopia, 2019
Treatment approach
|
Swallowed PZQ
|
X2
|
P-value
|
Yes (%)
|
No (%)
|
Unknown (%)
|
Vertical
|
3316(72.5)
|
1114(24.4)
|
141(3.1)
|
108.5
|
<0.001
|
Integrated
|
970(87.6)
|
121(10.9%)
|
17(1.5%)
|
Reported reasons for not swallowing PZQ among SAC in study districts, Ethiopia
The main reported reason for not swallowing PZQ against SCH among SAC was not attending school (n=394, 31.9%) (Figure: 3).
PZQ distribution sites reported by interviewee
Most of interviewed SACs in study districts reported that they had received PZQ treatment against SCH from school, 4044 (94.4%) (Fig: 4).
Factors statistically and significantly associated in multivariable logistic regression analysis
Multivariable logistic regression analysis showed that age of SAC, gender and school enrollment status of SAC was statistically and significantly associated with PZQ acceptance and swallowing. Students with age 10-14 years were 45% more likely to accept and swallow PZQ mass treatment compared with their counter parts (5-9 years); (AOR=1.45, 95% CI: 1.25, 1.69)
After stabilizing age and school attendance in the model constant and compare male SAC with female towards PZQ swallowing status, male SAC were 27% more likely to consume PZQ compared with their counter parts. (AOR= 1.27, 95% CI: 1.09, 1.47). Attendance to the school was statistically and significantly associated with PZQ swallowing against SCH. SAC enrolled to private or public school were almost 21 times more likely to swallow PZQ compared with SAC who didn’t; (AOR= 20.90, 95% CI: 17.41, 25.08). (Table: 5)
Model equation for the final fitted multivariable logistic regression analysis
(see Equation in the Supplementary Files)
Table 5: Factors statistically and significantly associated with PZQ swallowing status among
SAC in the sampled districts of Ethiopia, 2019
Variable
|
Category
|
Swallowed PZQ
|
COR
|
P-value
|
AOR
|
95% CI
|
Yes
N (%)
|
No
N (%)
|
Age
|
5-9 year
|
1,848 (43.1)
|
917 (65.83)
|
1
|
0.000
|
1
|
-
|
10-14 year
|
2,438 (56.9)
|
476 (34.17)
|
2.54
|
1.45
|
**(1.25, 1.69)
|
Gender
|
Female
|
2,092 (48.81)
|
761(54.63)
|
1
|
0.000
|
1
|
-
|
Male
|
2,194 (51.19)
|
632 (45.37)
|
1.26
|
1.27
|
*(1.09, 1.47)
|
School attendance
|
No
|
195 (4.55)
|
734 (52.69)
|
1
|
0.000
|
1
|
-
|
Yes
|
4,091(95.45)
|
659 (47.31)
|
23.37
|
20.90
|
**(17.41, 25.08)
|
Note: COR = Crude Odds ratio, AOR Adjusted Odds Ratio, CI = Confidence interval, *= P value < 0.05, ** < 0.01
Table 6: Final fitted model in multivariable logistic regression, 2019
To check the classifying or prediction performance of variables in the final fitted model, ROC curve was done. Based on the Hosmer and Lemeshow criteria, variables in the fitted model were predicting the acceptance or swallowing of PZQ by 77.62%.
This means 77.62% of PZQ swallowing among SAC was determined by age of SAC, gender of SAC and school attendance or enrollment status of SAC (Fig: 5).
Table 7: Test for instability of regression coefficient among variables in the fitted model