The ARMADILLO text message intervention to improve the sexual and reproductive health knowledge of adolescents in Peru: Results of a randomized controlled trial

BACKGROUND: The ARMADILLO Study determined whether adolescents able to access SRH information on-demand via SMS were better able to reject contraception-related myths and misconceptions as compared with adolescents receiving pushed SMS or no intervention. TRIAL DESIGN: This trial was an unblinded, three-arm, parallel-group, individual RCT with a 1:1:1 allocation. Trial registration: ISRCTN85156148. METHODS: This study was conducted in Lima, Peru among participants ages 13–17 years. Eligible participants were randomized into one of three arms: Arm 1: access to ARMADILLO’s SMS information on-demand; Arm 2 access to ARMADILLO SMS information pushed to their phone; Arm 3 control (no SMS). The intervention period lasted seven weeks. At baseline, endline, and follow-up (eight weeks following endline), participants were assessed on a variety of contraception-related myths and misconceptions. An index of myths-believed was generated. The primary outcome assessed the subject-specific change in the mean score between baseline and endline. Knowledge retention from endline to follow-up was also assessed, as was a ‘content exposure’ outcome, which assessed change in participants’ knowledge based on relevant SMS received. RESULTS: In total, 712 participants were randomized to the three arms: 659 completed an endline assessment and were included in the primary analysis. Arm 2 participants believed fewer myths at endline compared with control arm participants (estimated subject-specific mean difference of -3.69% [-6.17%, -1.21%], p = 0.004). There was no significant difference between participants in Arm 1 vs. the control Arm, or between participants in Arm 1 vs. Arm 2. A further decrease in myths believed between endline and follow-up (knowledge retention) was observed in all arms; however, there was no difference between arms. The content exposure analysis saw significant reductions in myths believed for Arm 1 (estimated subject-specific mean difference of -9.47% [-14.83%, -4.11%], p = .001) and Arm 2 (-5.93% [-8.57%, -3.29%], p < .001) as compared with the control arm; however Arm 1’s reduced sample size (n = 28) is a severe limitation. DISCUSSION: The ARMADILLO SMS content has a significant (but small) effect on participants’ contraception-related knowledge. Standalone, adolescent SRH digital health interventions may affect only modest change. Instead, digital is probably best used a complementary channel to expand the reach of existing validated SRH information and service programs.