Factors associated with unintended pregnancies among unmarried adolescents and the potential of using mobile money shops: A Cross-sectional study in Eastern Uganda

Study design and setting

The detailed study design has been published elsewhere [23] and was registered at J アパ根セ ‘the University Hospital Medical Information Network (UMIN000053332) in Japan. Here, we provide the fundamental parts of the study design for interpreting the study results. This was a cross-sectional study of unmarried adolescents aged 15-19 years in the Busoga region of Eastern Uganda. The center of the Busoga region is approximately 80 km from Kampala, the capital city. Two municipalities in the Busoga region with similar characteristics were selected: Iganga and Bugiri. In 2022, the population of Iganga municipality was 55,263 and that of Bugiri was 28,747, whose teenage population (10-17 years old) accounted for 20.9% and 19.3%, respectively. We selected 30 vendors from each municipality according to predefined study criteria. Owners of mobile money businesses (vendors) aged 18-30 years, who have operated for more than one year, and are willing to participate in the study were selected. Stand-alone shops were selected to secure the privacy of customers and commodities. The geographical distribution, diversity, population density, availabilities of health facilities and schools of each municipality were also considered. We recruited teenage customers from selected vendor sites. Recruitment ceased after reaching the desired sample size of approximately 20 per vendor.

Participants

The eligible study participants were unmarried girls/boys aged 15-19 years who used the selected vendors in the target municipalities and had no plans to move out of the municipalities during the study period. Adolescents aged 13-14 years were excluded because they tended to be less sexually active and used fewer vendors.

We applied a non-randomized prospective controlled design because vendor customers cannot be assigned to specific vendors, and they choose several vendors at their own convenience. We decided on a sample size of 1200 for the intervention study in the two municipalities using a formula based on the available data and previous studies, using a calculation method explained elsewhere [23]. Adolescents under 18 years of age who agreed to contact their parents/guardians during the screening process were recruited.

Data collection

A structured questionnaire was developed using research instruments based mainly on the 2016 Uganda Demographic and Health Survey [26] and other studies conducted in SSA [27-33]. We have added questions regarding social and physical opportunities, and those related to vendors’ usage [23]. We developed a questionnaire in English and translated it into the local language. Before finalizing the questionnaire, we conducted a series of pre-tests with 26 adolescents (10 boys and 16 girls) and modified them for easy answering. Although most adolescents used English daily, some participants were comfortable using their local language to discuss sensitive topics in SRHR. We employed data collectors who could manage the local language and had academic backgrounds in the social welfare or health fields. The questionnaire included five categories: (1) sociodemographic characteristics, (2) sexual behavior, (3) contraception knowledge and capability, (4) environmental factors preventing pregnancy, and (5) preference for contraceptive practices.

The questionnaire was administered face-to-face by trained data collectors using preprogrammed tablet computers [23]. Interviews were conducted in a private location near vendors’ or teenagers’ homes between February and March 2024. Data collection lasted between 30 and 40 min per interview.

Data analysis

The primary outcome was the experience of pregnancy experience of girls or impregnation of boys, regardless of sex. First, we described the sociodemographic characteristics of the participants and their sexual behaviors with the comparison of the two groups (pregnancy/impregnation [pregnant] versus never-pregnant) using the chi-square test. Second, we described the variables in each category using two-group comparisons. Third, multivariate logistic regression analysis was performed to identify the factors associated with the primary outcome. Variables that showed a statistically significant difference in bilateral regression were considered for multivariate logistic regression. In addition, we removed correlated variables and added the basic variables identified in a previous study, which accounted for 21 variables. Statistical significance was set at p < 0.05. Data analysis was conducted using STATA v.18.

Ethical consideration

Ethical approval was obtained from the Ethics Review Committee of Uganda Christian University (UCUREC-2023-710), the Uganda National Council for Science and Technology (SS2284ES), and the Ethics Review Committee of the JICA Ogata Sadako Research Institute for Peace and Development in Japan (JICA Ogata RI, reference: JICA [DI] 202312220002). Written informed consent was obtained from all adolescent participants and the parents/guardians of adolescents aged 15-17 years. Adolescents may feel uncomfortable participating in the study as the content includes their sexual behaviors. To mitigate their anxiety, we informed them that their participation was truly voluntary without any harm and that they could withdraw their participation at any time, even after the survey and collected information were strictly confidentially managed. To obtain consent from parents/guardians of adolescents under 18 years of age, highly experienced and skilled researchers visited the homes of expected participants and explained the study to the parents/guardians.

Active sexual behavior

Adolescents in SSA are sexually active [34, 35]. The Ugandan national average of percentage of those who had sexual intercourse among unmarried population aged 15-49 years were 59.6% for male and 45.6% of female in 2021 [9]. Our respondents showed more active sexual behaviors than the national trend, even limiting the younger population. The Ugandan national average age at first intercourse among adults aged 20-49 years was 18.3 years for males and 16.6 years for females [9]. Compared with the national average, our study participants had an early sexual initiation (15.4 years for both sexes; 5.0 years for males; 15.9 years for females). These trends were much lower than those in other SSA countries, for example, 17.0 years for females in Nigeria [27]. Furthermore, the pregnant group experienced the first sexual intercourse 0.44 years earlier than the other group in this study. This was in line with previous studies that revealed that earlier first intercourse was associated with higher unintended pregnancy rates in SSA [14, 34-36]. Our study found that the experience of pregnancy was 0.36 times lower among school students than among non-students, implying that pregnancy led to drop-out from school or out-school teens are more likely to face pregnancy risks. This supports that longer school years may reduce pregnancy risks, and extension of compulsory education years by the governmental strategy may partially effect to its purpose [14]. In addition, the Uganda government had issues several laws to reduce unintended girls pregnancy risk since 2000, such as the Unholy Law and the prohibition of marriage for girls under the age of 18 years [14]. However the data are limited to prove the effectiveness of these governmental compulsory measures [14]. Therefore, further research is needed to capture explicit adolescent needs to lower sexual activities nationwide and strengthen effective, reality-based policies [12, 14].

Contraceptive knowledge and practice

Non-practice of contraception is a critical determinant of unintended adolescent pregnancies in SSA [8, 36]. Most respondents in this study had a basic knowledge of contraceptive methods, which is similar to the national trend among sexually active unmarried both-sex teenagers (97.8% for boys and 99.4% for girls aged 15-19 years) [26]. However, the participants’ practical knowledge on how to use the methods was 16% points lower than their basic knowledge. In addition, only 9.1% of the adolescents had correct knowledge of the fertile period of the menstrual cycle, which was lower than the national average among girls aged 15-19 years (13.8%) in 2016 [26]. Our descriptive analysis revealed that adolescents with pregnancy experience tended to have more knowledge and access to contraception and support from parents/guardians and healthcare workers. Although this cross-sectional study could not identity causality, the results may be interpreted that teens those who experienced pregnancy became more serious about searching for information and mobilizing services after pregnancy. It can be said that Uganda’s SRHR education either at school, community, nor home does not provide preventative information of pregnancy and meet their real needs, although the Ministry of Education and Sports encourages the improvement of SRHR education in schools. In our study, the most common source of contraceptive information was schools; however, this was not the preferred source of SRHR information. A previous study concluded that insisting on abstinence in school education was ineffective in preventing teenage pregnancies [37]. Providing practical knowledge on sex, pregnancy, contraception, and sexually transmitted infections is crucial among adolescents at the scale, through home, school, and community [37-39]. Some studies emphasized that younger people were exposed to more unsafe sexual intercourse and were flexible in absorbing concepts related to gender, suggesting that sex education from an early age before sexual initiation was effective [36, 40-42]. A study on children aged 10-14 years in Uganda supported this suggestion [43]. Education content should include sex, pregnancy, delivery, sexually transmitted infections, other health risks, the efficacy of vulnerable populations, and future income potential [3, 7, 11, 32, 37, 38, 44]. The content should be designed according to the needs of different adolescent age groups [41, 45]. About half of those who had ever been pregnant were most likely to terminate their pregnancies. Since abortion for non-medical reasons is illegal in Uganda, statistics on abortion data were collected only through post-abortion care at health facilities and the actual situation is underreported. The previous qualitative studies captured that many girls have used traditional medicines or unsafe means to terminate unwanted pregnancies and received no after-abortion care [11, 12, 14]. Ignoring high-risk, unsafe self-abortion or necessary after-abortion care is likely to lead to maternal death and adverse maternal outcomes [12]. Sexual education emphasizing the risks of abortion and infertility is needed [11, 37].

Parental involvement

Parental involvement was an essential factor influencing adolescent unintended pregnancy in several studies in SSA countries [31, 32, 38]. In this study, "living with parents" was indicated to be a deterrent factor to sexual behavior. Studies in Ghana have reported a similar situation in which parental involvement reduced pregnancy occurrence [31, 38]. Conversely, parental interaction and parental contraceptive support were positively correlated with pregnancy experience, which may imply that parents’ substantive care and support increased after their children’s pregnancy experience, similar to the results of a Ghanaian study [31]. Parents’ neglect of their children and hesitancy to teach sex education because they did not know how to interact with them was also noted in a previous study in Uganda [11]. Parents’ motivation for sex education for their children, their own acquisition of correct SRHR knowledge, and their ability to interact with the children need to be improved [27, 32, 38]. Urgent development of educational program guidance for families (parents/guardians) and communities, including practical sex education content and skills to teach children in Uganda [11] is required.

Contraceptive accessibility in the community

The prevalence of unintended pregnancies is closely related to access to contraceptives [44]. Paid services at private clinics or pharmacies were hurdles for teenagers in Uganda and they tend to avoid public health facilities because they follow formal procedures, ask detailed questions about personal information and the purpose of contraceptive use, and, in some cases, may be known to their parents [11, 12]. This study found that motivation for contraceptive use was higher among those who had experienced pregnancy, and in fact, they mobilized a variety of contraceptive sources. These facts imply that teenagers who experienced pregnancies have strengthened their contraceptive capabilities. A Ghanaian study also pointed out that many girls who had experienced pregnancy developed competencies to deal with pregnancy prevention [31]. Our findings also indicated that easily accessible community-based distributors showed a preference for and a higher desire for future use among teenagers who had experienced pregnancy.

This study also found Approximately 70% of adolescents thought that the vendor was a suitable channel for receiving SRHR services. Furthermore, adolescents who have experienced pregnancy are more likely to prefer accessing vendor-based SRHR service. These facts suggest that vendors may be effective channels for providing SRHR services, especially for high risk population. Chandra-Mouli pointed out that interventions implemented in many programs such as facility-based interventions, including school education, health center activities, and youth centers with peer-educators in the community, have not been cost-effective or sustainable [19]. The Ugandan government has called for innovative approaches that utilize local mechanisms outside the health and education sectors [46, 47]. In response to the global community calling for innovative platforms to reach vulnerable populations nearly a decade ago [4, 17, 21], our results suggest that vendors in Uganda could be a cost-effective platform to reach adolescents and strengthen their competencies in accessing contraceptives and practice.