At baseline 6,055 individuals aged 15 to 22 years were screened of which 4,863 participated in the baseline survey, 4,421 were HIV negative and 4,389 chose to enrol into the study. As designed, 50.4% of those enrolled in the study were enrolled-in-education at the start of the study with ~ 50% in education and ~ 50% out-of-education within each of the study groups (Table 2). As designed across the study groups ~ 80% of the individuals were from rural enumeration areas (EAs) and ~ 20% were from urban EAs (Table 2). Figure 3 shows the cascade from screening to enrolment for the impact evaluation. The mean age of the sample was 18.22 years (standard deviation [SD]: ± 2.24). There were multiple significant differences (p-value < 0.05) when comparing urban versus rural individuals which included highest grade attained (fewer AGYW attaining grades 1–7 for urban and more AGYW attaining higher grades for urban), never having sexual intercourse (fewer urban than rural never having intercourse), condom use with a partner (more urban AGYW report using a condom than rural AGYW), ever giving birth (more rural than urban AGYW), last sexual partner circumcised (more urban AGYW than rural), heard of age mixing or transactional sex (both higher for rural than urban), and differences in household structure including female-headed households, number of siblings and currently married or in a union (Table 2). Additionally, there were significant differences (p-value < 0.05) when comparing AGYW enrolled in-education at the time of baseline collection and not enrolled-in-education at the time of baseline collection. These included age (older individuals being not enrolled in-education), school history (not enrolled-in-education individuals attained lower grades), sexual behaviours (not enrolled-in-education more likely of having had sexual intercourse, experienced sexual partner age mixing, used a condom with last partner, received money or gift in exchange for sexual intercourse, had given birth, sexual knowledge (not enrolled-in-education AGYW heard of HIV and had a higher level of knowledge of sexual practices than in-education AGYW), and household structure differences (Table 2).
Table 2
Characteristics of adolescent girls and young women at baseline
Characteristics | All Participants (N = 4863) | Urban (U) (N = 895) | Rural (R) (N = 3968) | U = R p-value | aEnrolled-in-Education (EIE) (N = 2337) | Not aEnrolled in-Education (NIE) (N = 2526) | EIE = NIE p-value |
Age in years (mean (SD)) | 18.22 (2.24) | 18.16 (2.28) | 18.24 (2.24) | | 16.94 (1.90) | 19.40 (1.86) | |
15 years (n (%)) | 779 (16.0%) | 143 (16.0%) | 636 (16.0%) | 0.16 | 693 (89.1%) | 86 (11.0%) | < 0.01 |
16 years (n (%)) | 594 (12.2%) | 134 (15.0%) | 460 (11.6%) | 473 (79.6%) | 121 (20.4%) |
17 years (n (%)) | 628 (12.9%) | 115 (12.9%) | 513 (12.9%) | 405 (64.5%) | 223 (35.5%) |
18 years (n (%)) | 612 (12.6%) | 104 (11.6%) | 508 (12.8%) | 281 (45.9%) | 331 (54.1%) |
19 years (n (%)) | 634 (13.0%) | 104 (11.6%) | 530 (13.4%) | 194 (30.6%) | 440 (69.4%) |
20 years (n (%)) | 636 (13.1%) | 108 (12.1%) | 528 (13.3%) | 143 (22.5%) | 493 (77.5%) |
21 years (n (%)) | 592 (12.2%) | 112 (12.5%) | 480 (12.1%) | 98 (16.6%) | 494 (83.5%) |
22 years (n (%)) | 388 (8.0%) | 75 (8.4%) | 313 (7.9%) | 50 (12.9%) | 338 (87.1%) |
School history | | | | | | | |
Highest grade attained | | | | | | | |
Grade 1–7 (n (%)) | 1156 (23.8%) | 182 (20.3%) | 974 (24.6%) | < 0.01 | 550 (47.6%) | 606 (52.4%) | < 0.01 |
Form 1–6 (n (%)) | 3581 (73.6%) | 668 (74.6%) | 2913 (73.4%) | 1700 (47.5%) | 1881 (52.5%) |
Year 1–5 (n (%)) | 126 (2.6%) | 45 (5.0%) | 81 (2.0%) | 87 (69.1%) | 39 (31.0%) |
bTime travelled to school > 30 minutes (n(%)) | 832 (36.4%) | 144 (28.5%) | 688 (38.6%) | | - | - | |
Sexual behaviours | | | | | | | |
Never had sexual intercourse (n (%)) | 2301 (47.3%) | 477 (53.3%) | 1824 (46.0%) | < 0.01 | 1785 (77.6%) | 516 (22.4%) | < 0.01 |
cSexual partner age mixing (n (%)) | 428 (20.3%) | 77 (22.5%) | 351 (19.8%) | 0.27 | 67 (15.6%) | 361 (84.4%) | 0.01 |
Sexual intercourse multiple partners in last 12 months (n (%)) | 25 (1.2%) | 7 (2.0%) | 18 (1.0%) | 0.11 | 7 (28.0%) | 18 (72.0%) | 0.32 |
Condom used with last partner (n (%)) | 1377 (65.1%) | 242 (70.6%) | 1135 (64.1%) | 0.02 | 334 (24.3%) | 1043 (75.7%) | < 0.01 |
Received money or gift in exchange for sexual intercourse (n (%)) | 586 (27.7%) | 88 (25.7%) | 498 (28.1%) | 0.35 | 99 (16.9%) | 487 (83.1%) | 0.03 |
Ever given birth (n (%)) | 1331 (52.0%) | 159 (38.0%) | 1172 (54.7%) | < 0.01 | 83 (6.2%) | 1248 (93.8%) | < 0.01 |
First pregnancy before the age of 17 (n (%)) | 505 (37.9%) | 63 (39.6%) | 442 (37.7%) | 0.64 | 34 (6.7%) | 471 (93.3%) | 0.56 |
Last sexual partner circumcised (n (%)) | 782 (42.3%) | 152 (52.6%) | 630 (40.4%) | < 0.01 | 163 (20.8%) | 619 (79.2%) | < 0.01 |
HIV and Sexual Practices Knowledge | | | | | | | |
Heard of AIDS (n (%)) | 4702 (96.7%) | 869 (97.1%) | 3833 (96.6%) | 0.45 | 2245 (47.8%) | 2457 (52.3%) | 0.02 |
Knows that a healthy looking person can have HIV (n (%)) | 4092 (84.2%) | 757 (84.6%) | 3335 (84.1%) | 0.69 | 1915 (46.8%) | 2177 (53.2%) | < 0.01 |
cHeard of age mixing (n (%)) | 3278 (67.6%) | 573 (64.7%) | 2705 (68.3%) | 0.04 | 1503 (45.9%) | 1775 (54.2%) | < 0.01 |
Heard of transactional sex (n (%)) | 3429 (71.0%) | 603 (68.1%) | 2826 (71.6%) | 0.04 | 1596 (46.5%) | 1833 (53.5%) | < 0.01 |
Household structure | | | | | | | |
Mother alive (n (%)) | 4057 (84.0%) | 759 (85.7%) | 3298 (83.6%) | 0.13 | 2002 (49.4%) | 2055 (50.7%) | < 0.01 |
Father alive (n (%)) | 3189 (67.0%) | 608 (69.1%) | 2581 (66.5%) | 0.15 | 1622 (50.9%) | 1567 (49.1%) | < 0.01 |
Female-headed household (n (%)) | 1592 (32.7%) | 336 (37.5%) | 1256 (31.7%) | < 0.01 | 739 (46.4%) | 853 (53.6%) | 0.11 |
Number of siblings (mean (SD)) | 1.46 (1.2) | 1.29 (1.1) | 1.49 (1.2) | < 0.01 | 1.54 (1.2) | 1.38 (1.2) | < 0.01 |
Currently married or in union (n (%)) | 115 (2.4%) | 11 (1.2%) | 104 (2.6%) | 0.01 | 0 (0.0%) | 115 (100.0%) | < 0.01 |
Table 2b footnotes:
aEnrolled in education: This variable corresponds to if an individual said she was enrolled in education at the time of the interview
bTime travelled to school: How long does it take the AGYW to travel from their house to school regardless of the time of day in minutes. Not answered by those not enrolled in education.
cAge mixing defined as having sexual intercourse with an individual 10 years or older than one’s self in the last 12 months
P-value are generated to show if there was a significant difference in Reponses between urban versus rural, and comparing enrolled in school and not enrolled in school; significant observation (p-value <0.05) are bolded
Responses of “Don’t know,” “Refuse to answer,” and “Not Applicable” are not included in this table. Missing values due to skip patterns are also not included in this table.
Column totals and percentages were calculated for locality, and row totals and percentages were calculated for in-school versus out-of-school |
Over half (52.7%) of the AGYW surveyed at baseline had their sexual debut. Additionally, 20.3% of AGYW who had sexual intercourse in the past 12 months, experienced age mixing (sexual intercourse in the last 12 months with an individual more than 10 years older than themselves), and 27.7% indicated that they had sexual intercourse in the exchange of money or gifts (Table 2). 32.7% of households were headed by females (Table 2). Very few (2.4%) AGYW were married or in a union at baseline (Table 2b).
The baseline prevalence of HIV, Trichomonas vaginalis, and syphilis among AGYW were 8.20% (397/4,840), 3.31% (150/4,533) and 0.17% (8/4,830) respectively (Table 3). The baseline prevalence of HIV among those enrolled-in-education was 3.7% (86/2,326) compared to 12.37% (311/2,514) among those not enrolled-in-education (Table 3). The baseline prevalence of Trichomonas vaginalis among those enrolled-in-education was 1.25% (27/2,153) compared to 5.17% (123/2,380) not enrolled-in-education. The baseline prevalence of syphilis among those enrolled-in-education was 0.09% (2/2,323) compared to 0.24% (6/2,507) not enrolled-in-education. 9.7% (473/4,863) of AGYW in the study sample had experienced sexual and gender-based violence. Factors in Table 4 that were significantly associated with increased risk of HIV infection at baseline were older age, not being enrolled in education. It should be noted that sexual partner age mixing was an extremely strong predictor of HIV with only 0.1% (2/1,685) of AGYW who did not engage in age mixing testing HIV positive while 65.5% (275/420) of AGYW who did engage in age mixing tested HIV positive.
Table 3
Prevalence of HIV, syphilis, and Trichomonas vaginalis among adolescent girls and young women at baseline
Laboratory-confirmed STI | Number (n) | Prevalence % (n/N) | In-education at Baseline | Out-of-education at Baseline | Rural | Urban |
HIV | 397 | 8.20 (397/4840)* | 3.70 (86/2326) | 12.37 (311/2514) | 7.92 (313/3,950) | 9.44 (84/890) |
Both HIV and Syphilis | 4 | 0.08 (4/4827) | 0.04 (1/2322) | 0.12 (3/2505) | 0.08 (3/3938) | 0.11 (1/889) |
Both HIV and Trichomonas vaginalis | 50 | 1.10 (50/4531) | 0.23 (5/2153) | 1.89 (45/2378) | 0.91 (34/3727) | 1.99 (16/804) |
Trichomonas vaginalis | 150 | 3.31 (150/4533) | 1.25 (27/2153) | 5.17 (123/2380) | 3.41 (127/3729) | 2.86 (23/804) |
Syphilis | 8 | 0.17 (8/4,830) | 0.09 (2/2323) | 0.24 (6/2507) | 0.15 (6/3941) | 0.23 (2/889) |
Both Syphilis and Trichomonas vaginalis | 3 | 0.07 (3/4532) | 0.05 (1/2152) | 0.08 (2/380) | 0.03 (1/3728) | 0.25 (2/804) |
*23 individuals without HIV test results; 330 individuals without Trichomonas vaginalis test results; 33 individuals without Syphilis test results |
Table 4
Factors associated with HIV infection in adolescent girls and young women at baseline
Factors | OR (95% CI) | p-value |
aAge in years | | |
15–18 years | 1 | |
19–22 years | 3.24 (2.60–4.04) | < 0.01 |
Locality | | |
Urban | 1 | |
Rural | 0.82 (0.64–1.05) | 0.12 |
School history | | |
Not benrolled in education | 1 | |
bEnrolled in education | 0.30 (0.24–0.37) | < 0.01 |
cHighest grade attained | | |
Grade 1–7 | 1 | |
Form 1–6 | 0.56 (0.45–0.70) | < 0.01 |
Year 1–5 | 0.61 (0.31–1.18) | 0.14 |
<30 minutes to travel to school | 1 | |
> 30 minutes to travel to school | 1.02 (0.67–1.56) | 0.92 |
Sexual behaviours | | |
Never had sexual intercourse | 1 | |
Had sexual intercourse | 4.18 (3.26–5.36) | < 0.01 |
cNo partner age mixing | 1 | |
cSexual partner age mixing | > 100.00 (346.5-3459.4) | < 0.01 |
No Sexual intercourse with multiple partners in last 12 months | 1 | |
Sexual intercourse with multiple partners in last 12 months | 2.04 (0.81–5.15) | 0.13 |
Condom not used with last partner if having multiple partners | 1 | |
Condom used with last sexual partner | 0.73 (0.57–0.94) | 0.02 |
Did not receive money of gift in exchange for sexual intercourse | 1 | |
Received money of gift in exchange for sexual intercourse | 1.08 (0.82–1.42) | 0.60 |
Never given birth | 1 | |
Ever given birth | 1.45 (1.15–1.83) | < 0.01 |
> 17 years old for first pregnancy | 1 | |
< 17 years old for first pregnancy | 1.31 (0.96–1.77) | 0.09 |
Last partner not circumcised | 1 | |
Last sexual partner circumcised | 0.98 (0.76–1.28) | 0.91 |
HIV and Sexual Practices Knowledge | | |
Has not heard of AIDS | 1 | |
Heard of AIDS | 1.08 (0.61–1.92) | 0.80 |
Knows that a healthy looking person can have HIV | 1 | |
Does not know that a healthy looking person can have HIV | 0.97 (0.74–1.28) | 0.84 |
Has not heard of age mixing | 1 | |
Heard of age mixing | 1.08 (0.87–1.34) | 0.49 |
Has not heard of transactional sex | 1 | |
Head of transactional sex | 1.33 (1.05–1.68) | 0.02 |
Syphilis and Trichomonas vaginalis status | | |
Does not have syphilis | 1 | |
Has syphilis | 10.84 (6.21–18.92) | < 0.01 |
Does not have Trichomonas vaginalis | 1 | |
Has Trichomonas vaginalis | 1.82 (1.59–2.09) | < 0.01 |
Household structure | | |
Mother alive | 1 | |
Mother deceased | 1.85 (1.46–2.34) | < 0.01 |
Father alive | 1 | |
Father deceased | 1.56 (0.27–1.92) | < 0.01 |
Male-headed household | 1 | |
Female-headed household | 1.09 (0.88–1.34) | 0.41 |
Number of siblings | | |
No siblings | 1 | |
1 sibling | 0.56 (0.42–0.74) | < 0.01 |
2 siblings | 0.63 (0.47–0.84) | < 0.01 |
> 2 siblings | 0.67 (0.52–0.86) | < 0.01 |
Not currently married or in union | 1 | |
Currently married or in union | 3.40 (2.18–5.30) | < 0.01 |
Table 4 footnotes:
OR: odds ratio
CI: confidence interval
Significant observation (p-value <0.05) are bolded
bEnrolled in education: This variable corresponds to if an individual said she was enrolled in education at the time of the interview
cAge mixing defined as having sexual intercourse with an individual 10 years or older than one’s self in the last 12 months
Responses of “Don’t know,” “Refuse to answer,” and “Not Applicable” are not included in this table. Missing values due to skip patterns are also not included in this table |
Factors in Table 4 that were significantly associated with decreased risk of HIV infection at baseline were being enrolled-in-education, higher grade of attainment, condom use with last partner and having siblings.