A total of 154 students were approached to participate in the study according to the eligibility criteria. Fourteen were excluded as they did not consent to be part of the study so 140 were enrolled in the study and all were successfully followed to the immediate interview. Ten students lost to follow up and 130 completed the assessment at the three months’ time making them the final participants for the study whose data is taken for analysis. Hence, the response rate was 92.9%. A Consolidated Standards of Reporting Trials flow chart showing subjects initially enrolled, intervened, followed up and included in the analyses in the quasi- experimental study is attached as supporting information (online supplemental material S2).
Personal and Family Characteristics
Table 2 shows the personal and family characteristics of the study participants. The minimum and maximum ages of the students were 14 and 18 years respectively, with a mean age of 15.36(SD = 1.02) years. The majority of the participants were orthodox (77.9%) and 7.9% were Catholic and Muslims each, and 2.1% were protestant. The percentage of students who live with their parents, mother only, father only, and others were 75.7%, 20%, 1.4%, and 2.9% respectively. Only 30% and 21.4% of the students always discuss sex-related issues with their mothers and fathers respectively, while 12.9%, and 36.4% respectively never discuss with their mother and father. The majority (67.1%) of the study participants had elder sisters, and 60% of them had elder brothers. Academic background of the students’ fathers, and mothers showed that 6.4% of fathers and 11.4% of mothers were illiterate, 13.6% of fathers and 29.3% of mothers were at elementary level, 28.6% of fathers, and 29.3% of mothers were at secondary level, and 46.4% of fathers and 30% of mothers were at post-secondary level of education. About 23% of them were not living with both biological parents.
Table 2
Personal and family characteristics of the study participants from secondary schools in Asmara, Eritrea, August to November 2019 (n = 130).
Variable name
|
Response Type
|
Number
|
Percentage
|
Age (Mean = 15.35, SD = 1.025)
|
|
|
13–15
|
82
|
63.1
|
|
16–18
|
48
|
36.9
|
Religion
|
|
|
|
Orthodox
|
100
|
76.8
|
|
Catholic
|
11
|
8.5
|
|
Moslem
|
8
|
6.2
|
|
Protestant
|
11
|
8.5
|
With whom she lives
|
|
|
|
Parents
|
98
|
75.4
|
|
Mother only
|
26
|
20.0
|
|
Father only
|
2
|
1.5
|
|
Other
|
4
|
3.1
|
Discuss sex issues with Father
|
|
|
Always
|
29
|
22.3
|
|
Frequently
|
9
|
6.9
|
|
Sometime
|
36
|
27.7
|
|
Rarely
|
8
|
6.2
|
|
Never
|
48
|
36.9
|
Discuss sex issues with Mother
|
|
|
Always
|
38
|
29.2
|
|
Frequently
|
15
|
11.5
|
|
Sometime
|
54
|
41.5
|
|
Rarely
|
6
|
4.6
|
|
Never
|
17
|
13.1
|
Presence of elder sister
|
|
|
|
Yes
|
86
|
66.2
|
|
No
|
44
|
33.8
|
Presence of elder brother
|
|
|
Yes
|
77
|
59.2
|
|
No
|
53
|
40.8
|
Have you ever gone to clubs
|
|
|
|
Yes
|
9
|
6.9
|
|
No
|
121
|
93.1
|
Difference in academic performance
|
|
Better
|
65
|
50.0
|
|
Less
|
20
|
15.4
|
|
No change
|
45
|
34.6
|
Educational level of father
|
|
|
Illiterate
|
8
|
6.2
|
|
Elementary
|
25
|
19.2
|
|
Secondary
|
38
|
29.2
|
|
Post high school
|
59
|
45.4
|
Educational level of mother
|
|
|
Illiterate
|
15
|
11.5
|
|
Elementary
|
37
|
28.5
|
|
Secondary
|
40
|
30.8
|
|
Post high school
|
38
|
29.2
|
Source of Knowledge on RH
The responses on the source of information regarding RH showed that the majority were life skill teachers at the school (57.9%). More than one fourth (27.9%) of the students also responded that they get information from their mothers, while 17.9% and 15.9% said from their friends and sisters respectively (Fig. 2).
Knowledge on RH
Out of the 13 items that assessed knowledge regarding RH, 11 items were correctly known by less than half of the students at pre-intervention (Table 3). The remaining two items: a girl stops growing after she makes sexual intercourse (80.0%), and EC should be taken with the consultation of a health care provider (66.2%); were correctly known by more than half of the students at pre-intervention. The item ‘when is EC taken?’ was the least (5.4%) correctly known, followed by the correct time at which a woman gets pregnant within the menstrual cycle (18.5%) before the intervention.
However, at immediate post, all the 13 items were correctly responded by more than 60% of the students. At immediate post, usage of emergency contraceptive after every sexual intercourse (58.5%) was the least correctly known followed by “selection of the suitable EC” (61.5%).
Moreover, the percentage of respondents who had correct knowledge on 3 items (out of the 13 items) returned to less than half, at 3 months follow-up. Four items were correctly responded by almost half, but the remaining were correctly known by more than 60% students.
Table 3
Percentage distribution of correct knowledge on items regarding RH (n = 130).
Knowledge on items regarding RH
|
PI
|
IP
|
3-MF
|
n (%)
|
n (%)
|
n (%)
|
Age at which menarche starts
|
60 (46.2)
|
90 (69.2)
|
70 (53.8)
|
The correct time at which a woman gets pregnant within the menstrual cycle
|
24 (18.5)
|
85 (65.4)
|
65 (50.0)
|
Whether a woman can get pregnant on the first time of sexual intercourse
|
63 (48.5)
|
90 (69.2)
|
82 (63.1)
|
Whether a girl stops growing after she makes sexual intercourse for the first time
|
104 (80.0)
|
126 (96.9)
|
121 (93.1)
|
Whether a woman is most likely to get pregnant if she has sexual intercourse half way between her periods
|
34 (26.2)
|
89 (68.5)
|
69 (53.1)
|
Whether a girl can take emergency contraceptives after every sexual intercourse
|
33 (25.4)
|
76 (58.5)
|
58 (44.6)
|
What method of EC have they heard?
|
39 (30.0)
|
100 (76.9)
|
71 (54.6)
|
Which method of contraceptives are suitable for adolescents?
|
33 (25.4)
|
80 (61.5)
|
53 (40.8)
|
The female should carry EC pill with her, so that she can use it after sex
|
44 (33.8)
|
98 (75.4)
|
84 (64.6)
|
EC can protect from STI
|
45 (34.6)
|
90 (69.2)
|
86 (66.2)
|
Use of EC will harm pregnancy
|
34 (26.2)
|
112 (86.2)
|
96 (73.8)
|
EC should be taken with knowledge and consultation of a health care provider
|
86 (66.2)
|
109 (83.8)
|
92 (70.8)
|
When is EC taken?
|
7 (5.4)
|
84(64.6)
|
52(40.0)
|
PI = Pre-intervention, IP = Immediate Post, 3-MF = Three months follow-up. |
Effect of Intervention through time on knowledge regarding RH
The mean (SD) scores of knowledge on RH at time 1(pre-intervention), Time 2 (immediate post) and Time 3(3-months follow-up) were 6.57 (2.08), 9.26 (2.14), and 8.46 (2.69) out of 13 respectively (Table 4). Mauchly’s test indicated that there was an evidence of non-sphericity (Mauchly’s W = 0.931, p = 0.010). Hence, significant effect for time (Wilks’ Lambda = 0.508, F (2, 128) = 61.95, p < 0.001), was observed using multivariate ANOVA.
Table 4
Level of knowledge regarding RH and the effect of educational intervention through time (n = 130).
Time Period
|
N
|
Mean (SD)
|
Effect with time
|
Time 1 (Pre-intervention)
|
130
|
6.57 (2.08)
|
Wilk’s λ = 0.508,
F(2,128) = 61.95,
p < 0.001
|
Time 2 (Immediate Post)
|
130
|
9.26 (2.14)
|
Time 3 (3-month follow-up)
|
130
|
8.46 (2.69)
|
Pairwise comparisons of the knowledge scores at the three time periods were also done using Bonferroni post-hoc test (Table 5). The test showed that knowledge scores were significantly higher at immediate post intervention than pre-intervention (p < 0.001), 3-months follow-up than pre-intervention (p < 0.001), and immediate post intervention than 3 months later (p = 0.005). The trend of knowledge scores regarding RH was observed to be quadratic across the three time periods (p < 0.001).
Table 5
Pair-wise comparison of knowledge scores regarding RH at pre-intervention, immediate post, and 3-months follow-up (n = 130).
Post-hoc Comparison
|
MD (95% CI)
|
p-value
|
p-value
trend
|
Immediate Post and pre-intervention
|
2.69 (2.11, 3.28)
|
< 0.001
|
< 0.001
|
3-months follow-up and pre-intervention
|
1.89 (1.18, 2.61)
|
< 0.001
|
Quadratic
|
Immediate Post and 3-months follow-up
|
-0.80 (-1.40, -0.20)
|
0.005
|
|
Positive attitude on RH
From 23 items proposed to assess the attitude towards RH, less than half of the study participants were found to have positive attitude in 11 items during pre-intervention. At pre-intervention, the percentage of female students who had positive attitude towards avoiding sex before they become engaged and sex only for one time is not good were 78.5% and 70.0% respectively. On the other hand, only one fourth of the female students had a positive attitude on how to use a condom properly (being in favour of abstinence).
At immediate post intervention, less than half of the study participants were found to have a positive attitude in 7 items. Friends never think of having an abortion (32.3%), and friends believe that you should be in love before you have sex with someone (30.8%) were the least improved items at immediate after intervention. Avoiding sex before they become engaged (78.5%), and sex for only one time should be avoided (76.2%) were the two items in which the students had relatively higher positive attitude. At 3-months follow-up, less than half of the study participants were found to have positive attitude in 11 items out of the 23 items.
Table 6
Percentage distribution of positive attitude on RH (n = 130).
Attitude on …
|
PI
|
IP
|
3-MF
|
n (%)
|
n (%)
|
n (%)
|
Sex education should be given to teenagers
|
83 (63.8)
|
95(73.1)
|
82 (63.1)
|
Alright for unmarried boys and girls to date
|
41 (31.5)
|
52 (40.0)
|
45 (34.6)
|
Alright for unmarried boys and girls to kiss, hug and touch each other
|
86 (66.2)
|
89 (68.5)
|
80 (61.5)
|
It is wrong for unmarried boys and girls to have sexual intercourse even if they love each other
|
87 (66.9)
|
98 (75.4)
|
94 (72.3)
|
A boy has to force a girl to have sex if he loves her
|
76 (58.5)
|
94 (72.3)
|
83 (63.8)
|
A boy will not respect a girl who has sex with him
|
43 (33.1)
|
51 (39.2)
|
48 (36.9)
|
Most girls who have sex before marriage regret it afterwards
|
88 (67.7)
|
91 (70.0)
|
55 (42.3)
|
A boy and a girl should avoid sex before they become engaged
|
102 (78.5)
|
115 (88.5)
|
82 (63.1)
|
Girls should remain virgin until they marry
|
62 (47.7)
|
70 (53.5)
|
57 (43.8)
|
Sex only for one time is not good
|
99 (76.2)
|
111 (85.4)
|
100 (76.9)
|
I am confident that I can insist on use of condom every time I have sex
|
66 (50.8)
|
75 (57.7)
|
72(55.4)
|
I would not support for a girl to have an abortion
|
49 (37.7)
|
72 (55.4)
|
55 (42.3)
|
A girl should be in love before having sex with a boy
|
47 (36.2)
|
46 (35.4)
|
46 (35.4)
|
Know how to use a condom properly (Being in favour of abstinence)
|
23 (17.7)
|
30 (23.1)
|
56 (43.1)
|
Friends never think of having an abortion
|
42 (32.3)
|
55 (42.3)
|
47 (40.0)
|
Friends believe that you should be in love before you have sex with someone (Being in favor of abstinence)
|
40 (30.8)
|
54 (41.5)
|
47 (36.2)
|
No to have sex without condom (Being in favour of abstinence)
|
57 (43.8)
|
71 (54.6)
|
60 (46.2)
|
Sex for only one time should be avoided
|
91 (70.0)
|
112 (86.2)
|
110 (84.6)
|
EC should be available at health facilities
|
73 (56.2)
|
115 (88.5)
|
101 (77.7)
|
It is not alright for girls to have sex even if they use methods to stop pregnancy
|
37 (28.5)
|
44 (33.1)
|
70 (53.8)
|
It is mainly the woman’s responsibility to use contraceptive regularly
|
37 (28.5)
|
60 (46.2)
|
49 (37.7)
|
They are not concerned about confidentiality
|
79 (60.8)
|
101 (77.7)
|
90 (69.2)
|
The girl shouldn’t carry the EC in her bag
|
68 (52.3)
|
112 (86.2)
|
104 (80.0)
|
PI = Pre-intervention, IP = Immediate Post, 3-MF = Three months follow-up. |
Effect of Intervention through time on Attitude towards RH
The mean (SD) scores of attitude towards RH at time 1(pre-intervention), Time 2 (immediate post) and Time 3 (three month follow-up) were 48.98 (4.68), 51.03 (4.04), and 49.65 (4.01)out of 69 respectively (Table 7).
Mauchly’s test indicated that there was an evidence of non-sphericity (Mauchly’s W = 0.951, p = 0.040). Multivariate ANOVA showed significant effect for time (Wilks’ Lambda = 0.827, F (2, 128) = 13.42, p < 0.001).
Table 7
Level of attitude regarding RH and the effect of educational intervention through time (n = 130).
Time Period
|
N
|
Mean (SD)
|
Effect with time
|
Time 1 (Pre-intervention)
|
130
|
48.98 (4.68)
|
Wilk’s λ = 0.827,
F(2,128) = 13.42,
p < 0.001
|
Time 2 (Immediate Post)
|
130
|
51.03 (4.04)
|
Time 3 (3-month follow-up)
|
130
|
49.65 (4.01)
|
Table 8 shows the pairwise comparison of attitude scores regarding RH. Bonferroni post-hoc test showed that attitude scores was significantly higher at immediate post intervention than pre-intervention (p < 0.001) and immediate post intervention than 3 months follow-up (p = 0.001). However, no significant difference in attitude scores was found between 3 months follow-up and pre-intervention (p = 0.395). The trend of attitude towards RH and pregnancy scores was observed to be quadratic across the three time periods (p < 0.001).
Table 8
Pair-wise comparison of attitude scores regarding RH at pre-intervention, immediate post, and 3-months follow-up (n = 130)
Post-hoc Comparison
|
MD (95% CI)
|
p-value
|
p-value trend
|
Immediate post and pre-intervention
|
2.05 (1.01, 3.10)
|
< 0.001
|
< 0.001
|
3 months later and pre-intervention
|
0.68 (-0.41, 1.76)
|
0.395
|
Quadratic
|
Immediate post and 3 months later
|
1.38 (0.48, 2.27
|
0.001
|
|
Effectiveness of Educational Intervention on Knowledge regarding RH by demographic variables
Analysis using mixed between-within subjects analysis of variance was conducted to assess the impact of the educational intervention on the scores of knowledge about RH across three time periods by age, religion, discussion of sexual issues with father, discussion of sexual issues with mother, presence of elder sister, educational level of father and educational level of mother (Table 9). The result revealed no significant interaction between age group and time (Wilk’s Lambda = 0.991, p = 0.560), discussion of sexual issues with father (Wilk’s Lambda = 0.981, p = 0.662), mother (Wilk’s Lambda = 0.975, p = 0.529), presence of elder sister (Wilk’s Lambda = 0.978, p = 0.242), educational level of father (Wilk’s Lambda = 0.986, p = 0.782) and mother (Wilk’s Lambda = 0.964, p = 0.274). However, there was significant interaction between religion and time (Wilk’s Lambda = 0.950, p = 0.040).
Table 9
Effectiveness of educational intervention on knowledge of RH by background characteristics (n = 130).
Characteristics
|
PI
|
IP
|
3-MF
|
Box M
|
Wilk's Lambda
|
p-value
|
M (SD)
|
M (SD)
|
M (SD)
|
Age
|
|
|
|
|
|
|
|
13 to 15
|
6.69 (2.09)
|
9.20 (2.08)
|
8.37 (2.52)
|
0.426
|
0.991
|
0.560
|
|
16 to 18
|
6.37 (2.08)
|
9.37 (2.25)
|
8.61 (2.98)
|
|
|
|
Religion
|
|
|
|
|
|
|
|
Christian
|
6.46 (2.05)
|
9.17 (2.17)
|
8.55 (2.66)
|
0.364
|
0.950
|
0.040*
|
|
Muslim
|
7.73 (2.15)
|
10.27 (1.49)
|
7.55 (3.01)
|
|
|
|
Sex Issues Discussion with Father
|
|
|
|
|
|
Always/Frequently
|
6.05 (1.80)
|
9.00 (2.29)
|
7.97 (2.66)
|
0.628
|
0.981
|
0.662
|
|
Sometimes
|
6.89 (2.04)
|
9.00 (1.90)
|
8.47 (2.22)
|
|
|
|
|
Rarely/Never
|
6.71 (2.25)
|
9.61 (2.16)
|
8.79 (2.97)
|
|
|
|
Sex Issues Discussion with Mother
|
|
|
|
|
|
Always/Frequently
|
6.74 (2.04)
|
9.04 (2.25)
|
8.06 (2.60)
|
0.575
|
0.975
|
0.529
|
|
Sometimes
|
6.69 (2.09)
|
9.65 (1.81)
|
8.85 (2.63)
|
|
|
|
|
Rarely/Never
|
5.91 (2.13)
|
9.87 (2.49)
|
8.48 (2.99)
|
|
|
|
Presence of elder sister
|
|
|
|
|
|
|
Yes
|
6.62 (2.18)
|
9.34 (2.18)
|
8.24 (2.56)
|
0.497
|
0.978
|
0.242
|
|
No
|
6.48 (1.90)
|
9.11 (2.07)
|
8.89 (2.91)
|
|
|
|
Father's Educational Level
|
|
|
|
|
|
|
Junior or below
|
6.58 (2.26)
|
9.21 (2.13)
|
8.12 (2.09)
|
0.064
|
0.986
|
0.782
|
|
Secondary
|
6.32 (2.21)
|
9.21 (2.43)
|
8.13 (2.69)
|
|
|
|
|
Post-secondary
|
6.73 (1.91)
|
9.26 (2.14)
|
8.86 (2.96)
|
|
|
|
Mother's Educational Level
|
|
|
|
|
|
|
Junior or below
|
6.54 (2.00)
|
9.04 (2.43)
|
8.27 (2.08)
|
0.090
|
0.964
|
0.274
|
|
Secondary
|
6.30 (2.30)
|
9.78 (1.67)
|
8.83 (2..99)
|
|
|
|
|
Post-secondary
|
6.89 (1.96)
|
9.03 (2.10)
|
8.34 (3.11)
|
|
|
|
PI = Pre-intervention, IP = Immediate Post, 3-MF = Three months follow-up. |
Effectiveness of Educational Intervention on attitude on RH by demographic variables
Analysis using mixed between-within subjects analysis of variance was conducted to assess the impact of the educational intervention on the scores of attitude on RH across three time periods by age, religion, discussion of sexual issues with father, discussion of sexual issues with mother, presence of elder sister, educational level of father and educational level of mother (Table 10). The result revealed no significant interaction between age group and time (Wilk’s Lambda = 0.996, p = 0.772), religion and time (Wilk’s Lambda = 0.982, p = 0.308), discussion of sexual issues with father (Wilk’s Lambda = 0.966, p = 0.357), discussion of sexual issues with mother (Wilk’s Lambda = 0.973, p = 0.489), presence of elder sister (Wilk’s Lambda = 0.099, p = 0.133), and educational level of father (Wilk’s Lambda = 0.940, p = 0.099). However, border significance of interaction was observed between educational level of mother and time (Wilk’s Lambda = 0.927, p = 0.049).
Table 10
Effectiveness of educational intervention on attitude of RH by background characteristics (n = 130).
Characteristics
|
PI
|
IP
|
3-MF
|
Box M
|
Wilk's Lambda
|
p-value
|
M (SD)
|
M (SD)
|
M (SD)
|
Age
|
|
|
|
|
|
|
|
13 to 15
|
49.36 (4.71)
|
51.44 (3.90)
|
50.25 (3.66)
|
0.515
|
0.996
|
0.772
|
|
16 to 18
|
48.35 (4.61)
|
50.35 (4.21)
|
48.67 (4.40)
|
|
|
|
Religion
|
|
|
|
|
|
|
|
Christian
|
48.76 (4.63)
|
51.01 (4.12)
|
49.62 (4.13)
|
0.500
|
0.982
|
0.308
|
|
Muslim
|
51.27 (4.84)
|
51.27 (3.23)
|
50.00 (2.37)
|
|
|
|
Sex Issues Discussion with Father
|
|
|
|
|
|
Always/Frequently
|
47.21 (3.89)
|
49.84 (4.08)
|
48.84 (3.85)
|
0.734
|
0.966
|
0.357
|
|
Sometimes
|
48.86 (5.31)
|
51.47 (4.11)
|
50.17 (4.15)
|
|
|
|
|
Rarely/Never
|
50.25 (4.40)
|
51.55 (3.87)
|
49.88 (4.01)
|
|
|
|
Sex Issues Discussion with Mother
|
|
|
|
|
|
Always/Frequently
|
48.68 (4.13)
|
49.96 (4.10)
|
49.25 (4.09)
|
0.204
|
0.973
|
0.489
|
|
Sometimes
|
48.93 (5.26)
|
51.61 (3.91)
|
49.87 (3.93)
|
|
|
|
|
Rarely/Never
|
49.78 (4.55)
|
52.03 (4.04)
|
50.09 (4.10)
|
|
|
|
Presence of elder sister
|
|
|
|
|
|
|
Yes
|
48.81 (5.02)
|
51.29 (4.16)
|
49.42 (4.19)
|
0.099
|
0.969
|
0.133
|
|
No
|
49.30 (3.97)
|
50.52 (3.79)
|
50.11 (3.64)
|
|
|
|
Father's Educational Level
|
|
|
|
|
|
|
Junior or below
|
49.12 (4.77)
|
50.06 (3.94)
|
50.15 (3.75)
|
0.479
|
0.940
|
0.099
|
|
Secondary
|
47.97 (4.45)
|
51.26 (4.68)
|
49.08 (3.74)
|
|
|
|
|
Post-secondary
|
49.54 (4.74)
|
51.42 (3.61)
|
49.75 (4.32)
|
|
|
|
Mother's Educational Level
|
|
|
|
|
|
|
Junior or below
|
48.54 (4.48)
|
50.48 (4.43)
|
49.40 (3.88)
|
0.314
|
0.927
|
0.049*
|
|
Secondary
|
48.58 (3.95)
|
51.85 (4.15)
|
50.80 (3.86)
|
|
|
|
|
Post-secondary
|
50.00 (5.54)
|
50.92 (3.25)
|
48.79 (4.15)
|
|
|
|
PI = Pre-intervention, IP = Immediate Post, 3-MF = Three months follow-up. |