Preventing Teenage Pregnancy for Girls in Bangladesh

The main treatment intervention at the village level includes the following components: (1) Girls’ (Kishori) club: Monthly learning and counseling sessions with target girls in the village led by trained community health workers called “field motivators” on child marriage and reproductive health. Participants receive a reproductive health knowledge booklet prepared by the public health experts. At each monthly club session, field motivators record attendance and administer mini quizzes to assess the girls’ learning levels. (2) Family planning support: Family Planning Kit are provided to girls with referrals distributed personally. Girls who have migrated to their husband’s village will be followed up by phone and encouraged to connect with community health workers in their new village.

We also conduct an auxiliary light-touch intervention, which provides information-treated households with correct information about community beliefs, following the approach of Bursztyn et al. (2020). Randomization occurs at the village level. Specifically, we include questions in the baseline survey to assess the beliefs of girls and their parents about marriage and pregnancy, calculate the village averages, and provide households in the information-treated villages with accurate information about what others in the community believe.

2024-02-06

2025-02-05

Reproductive health knowledge, Use of contraceptives, Age of first marriage and first pregnancy, Belief about social norms

The variable measuring reproductive health knowledge will be constructed based on the principal commponent analysis of the correct/affirmative answers to the following questions (note that the first number, e.g., D8, corresponds to the question numbers in the questionnaire. See "Documents" section in this registry for the full questionnaire):
Questionnaire Section 5: Questions to daughter
D8. On Average, how frequently menstrual regulation happens for an adult girl?
D9. When during a menstrual cycle a girl is more likely to get pregnant, if she has sexual relations?
D10. Can you please tell me, what are the physical changes that a girl experiences when transitions into adulthood? (Multiple answers possible, answer all that apply)
D11. If a girl gets pregnant in her teenage years (before the age of 19), what are the potential issues or health risks she might have to face? (multiple answers possible, answer all that apply)
D12. Do you know the places and persons you can go if you have any reproductive health problems or need for consultation?
D13. Who do you discuss with if you face any reproductive health related problem, including puberty and menstruation?
D14. Do you know how a girl gets pregnant?
D15. Have you ever heard of following contraceptives to avoid pregnancy? (multiple answers, answer all that apply)

Use of contraceptives is binary variables for each of the following contraceptive methods, pills, condoms, injection, withdrawals, and others.

Marriage and pregnancy timing will be measured by the exact day, month, year of marriage and pregnancy. Age at first marriage and pregnancy is calculated using the exact day, month, year of girls' birthday.

The variable measuring belief about social norms will be constructed based on the principal commponent analysis of the answers conforming to the conservative/traditional norms of the following questions:
Questionnaire Section 5: Questions to daughter
D27. Say there are 100 families in your village/community. If you had to guess, how many families among 100 in the village/community would agree with the same statement as stated above in D26 with you? We will match your response with the answers collected from this village/community and if your answer matches correctly with the community response, you will be rewarded. (note D26. Do you agree with the following statement? In my opinion girls should not get pregnant immediately after marriage. They should follow family planning to delay pregnancy by at least 1-2 years after marriage to adjust with new life and should become physically and mentally mature to take the childbearing role.)
D30. Say there are 100 families in your village/community. If you had to guess, how many families among 100 in the village/community would agree with the same statement as stated above in D29 with you? We will match your response with the answers collected from this village/community and if your answer matches correctly with the community response, you will be rewarded. (note D29. Do you agree with the following statement? In my opinion girls should continue education after marriage.)
D34. Say there are 100 families in your village/community. If you had to guess, how many families among 100 in the village/community would agree with the same statement as stated above in D32 with you? We will match your response with the answers collected from this village/community and if your answer matches correctly with the community response, you will be rewarded. (note D32. Do you agree with the following statement? In my opinion girls should not get married before 18 years of age)

School continuation, Empowerment, decision-making and autonomy, Intimate partner violence (IPV)

School continuation is binary variable. Additionally, we collect years of schooling for those who dropped out of school.

The variable measuring empowerment, decesion-making and autonomy will be constructed based on the principal commponent analysis of the correct/affirmative answers to the following questions (note that the first number, e.g., D1, corresponds to the question numbers in the questionnaire. See "Documents" section in this registry for the full questionnaire):
Questionnaire Section 5: Questions to daughter
D1. Please tell us your educational aspiration.
D2. Please tell us your job aspiration in the future. Do you want to work for pay?
D3. At which age do you like to marry? Ideal age of marriage? State in years.
D4. Do you know what is the legal minimum age at marriage for girls? State in years.
D6. Do you have any say (decision making) on your own marriage?
D18. Do you think it important to get pregnant right after the marriage? -> The answer options relevant to empowerment are 2= She is going to have a problem with her husband and in-laws if she does not get pregnant soon after the marriage. 3= Late pregnancy brings a lot of physical and family related complicacy
D20. Do you think your future in-laws will pressurize to get pregnant right after the marriage?
D21. Do you think your future husband will pressurize to get pregnant right after the marriage?
D22. Do you think it a good idea for unmarried girls to have practical reproductive health education and knowledge before marriage?
D24. Do you think that you have the correct reproductive health education and knowledge?
Questionnaire Section 4: Attitudes 4-2 Decision making, Please answer the % of involvement for each -> DMA_d captures the level of daugher's involvement to the household decision making.
DMA1. What to cook on the daily basis?
DMA2. Whether to buy an expensive item such as TV or fridge?
DMA7. Whether your daughter should work?
DMA8. If your daughter takes a job, which job?
DMA9. To whom your children should marry?
DMA10. At what age your children should marry?

IPV is the binary variable for each type of IPV, i.e., verbal, physical (pushed, slapped, hit), and sexual.

The current study plans to conduct a cluster randomized controlled trial (cRCT) in the Gaibandha district, one of the most child marriage prevalent areas in rural Bangladesh. To achieve minimum statistical power of 0.8 in this cRCT, we have included a total of 120 villages, i.e., 60 treated, and 60 controlled. We randomly select 120 villages from those with high rates of child marriage and secondary school dropout based on the Participatory Rural Appraisal/Census. A condition for eligible villages is that no existing NGO activity related to this project is present.

The target of the intervention is adolescent girls aged 13–17 at baseline. We conduct listing surveys from September to October 2023, aiming to identify all eligible households in each of the 120 survey villages. Eligible households are defined as those with at least one unmarried girl aged 13–17. Based on the listing survey, we randomly select 10 households in each of the 120 villages to conduct baseline surveys from December 2023 to January 2024. Once the baseline is completed, half of the clusters (60 villages) will be randomly allocated to the treatment group, while the other half will remain as an experimental control (status quo) group. The intervention is planned from February 2024 to January 2025, followed by midline and endline surveys.

The main treatment intervention at the village level includes the components: (1) Girls’ (Kishori) club, (2) Family planning support. We also conduct an auxiliary light-touch intervention, which provides information-treated households with correct information about community beliefs. Randomization occurs at the village level, with half of the treated and controlled villages in the main intervention becoming information treatment villages.

Not available

Randomization done in office by a computer at the village level. Each of the 120 villages is assigend either treated and controlled groups in the main intervention (monthly girls' club intervention). Half of the treated and controlled villages of the main intervention are light-touch information-treated groups.

Village-level randomization.

Yes

120 villages

1200 girls (120 households, one girl in each household)

60 treated villages and 60 controlled villages in the main intervention, i.e., monthly girls' club intervention
60 light-touch information-treated vilalges (30 treated and 30 controlled of the mian intervention) and 60 information-controlled villages.

The sample size is 1200. We calculate the minimum detectable effect size (MDES) to ensure 0.8 statistical power. According to the latest DHS Bangladesh 2017–18, the rate of child marriage for women aged 20–24 is 70% in the Rangpur region, where our study site, Gaibandha district, is located. The DHS records the age of the first birth for married women, and we assume that the woman was pregnant one year before the age of first birth. Since the DHS records pregnancy only for married girls and out-of-wedlock pregnancy is unrealistic in Bangladesh, the rate of pregnancy below age 18 for married girls aged 13–17 is around 54%. Using the same data, we calculate the intracluster correlation (ρ=0.332). We calculated the MDES using the STATA command clustersampsi, with the control mean (without treatment) of 0.54, and control standard deviation of 0.45. The number of individuals per cluster is 10, with 60 clusters per arm. With a type I error at the 0.05 level, we estimate a minimum detectable size of 18.5%, i.e., the treatment mean of 0.44 or 44% girls below age 18 are pregnant, and treatment standard deviation of 0.47.