Busting contraception myths and misconceptions among youth in Kwale County, Kenya: results of a digital health randomised control trial

Objectives The objective of this randomised controlled trial in Kenya was to assess the effect of delivering sexual and reproductive health (SRH) information via text message to young people on their ability to reject contraception-related myths and misconceptions. Design and setting A three-arm, unblinded randomised controlled trial with a ratio of 1:1:1 in Kwale County, Kenya. Participants and interventions A total of 740 youth aged 18–24 years were randomised. Intervention arm participants could access informational SRH text messages on-demand. Contact arm participants received once weekly texts instructing them to study on an SRH topic on their own. Control arm participants received standard care. The intervention period was 7 weeks. Primary outcome We assessed change myths believed at baseline and endline using an index of 10 contraception-related myths. We assessed change across arms using difference of difference analysis. Results Across arms, <5% of participants did not have any formal education, <10% were living alone, about 50% were single and >80% had never given birth. Between baseline and endline, there was a statistically significant drop in the average absolute number of myths and misconceptions believed by intervention arm (11.1%, 95% CI 17.1% to 5.2%), contact arm (14.4%, 95% CI 20.5% to 8.4%) and control arm (11.3%, 95% CI 17.4% to 5.2%) participants. However, we observed no statistically significant difference in the magnitude of change across arms. Conclusions We are unable to conclusively state that the text message intervention was better than text message ‘contact’ or no intervention at all. Digital health likely has potential for improving SRH-related outcomes when used as part of multifaceted interventions. Additional studies with physical and geographical separation of different arms is warranted. Trial registration number ISRCTN85156148.

2. Please reconsider use of the terms 'myths and misconceptions'. Side-effects experienced from contraception methods can be real and not always be attributed to myths and misconceptions. Therefore please consider clarification regarding what you mean by a myth and a misconception and how this differs from a real experienced side-effect such as menstrual changes, feeling of body heat, skin changes, delayed return to fertility (with injectable) or rare serious health events (such as thromboembolism). This comment applies to several sections of the manuscript e.g. introduction (para 2), box 1 (please define what is meant by 'health problems'?), and discussion. Sorry, this is just my opinion but may not be shared by others and happy for the authors to push back on this. Some papers have discussed these issues e.g. Side effects and the need for secrecy: characterising discontinuation of modern contraception and its causes in Ethiopia using mixed methods and Menstrual Bleeding Changes Are NORMAL: Proposed Counselling Tool to Address Common Reasons for Non-Use and Discontinuation of Contraception. 2. Please reconsider use of the terms 'myths and misconceptions'. Side-effects experienced from contraception methods can be real and not always be attributed to myths and misconceptions. Therefore please consider clarification regarding what you mean by a myth and a misconception and how this differs from a real experienced side-effect such as menstrual changes, feeling of body heat, skin changes, delayed return to fertility (with injectable) or rare serious health events (such as thromboembolism). This comment applies to several sections of the manuscript e.g. introduction (para 2), box 1 (please define what is meant by 'health problems'?), and discussion. Sorry, this is just my opinion but may not be shared by others and happy for the authors to push back on this. Some papers have discussed these issues e.g. Side effects and the need for secrecy: characterising discontinuation of modern contraception and its causes in Ethiopia using mixed methods and Menstrual Bleeding Changes For the reviewer's ease and interest, we provide additional information below: In our own identification of this study's index of 10 myths and misconceptions, we started with a list of region-specific myths/misconceptions (identified from the literature) and then refined these heavily based on formative work done in the study area prior to the start of the RCT. Here, young men and women were asked to reflect on some short stories featuring couples their age who were considering starting contraception but were 'nervous about what they had heard from friends'. They were then asked to share what these couples may have heard -this reflects the broad nature (e.g. the 'health problems') of one in the list.
The study group refined the list to feature those identified as the most salient based on FGDs. We acknowledge that reductions in libido (specific to certain methods like injectables) and injectable-specific return to fertility delay that the reviewer mentions can still represent actual side effects for some women for some methods. We contend that the extrapolation of these method-specific experiences to contraception in general is part of what turns these into to persistent and widespread misconceptions about all contraception. An unfortunate result is the high discontinuation of contraception altogether because of side-effects experienced with one method, rather than seeking to find a better fit.
3. Few similar studies of digital health interventions for contraception are cited apart from McCarthy et al.
Most studies in the background are about STIs/HIV yet the intervention is about contraception. Quite a few studies exist so please consider include some. And then discuss findings of your studies against those studies in the discussion.
We have added references to two additional papers in the introduction but note that for the introductory paragraph we have endeavoured to reflect the real impetus that resulted in the ARMADILLO study's creationthe availability of thin but promising data in relevant areas of health, but none at the time specific to digital health interventions targeted at improving contraception outcomes among young people.
We have also added additional, more recent citations to the third paragraph in the discussion where we note our null findings (including results from two more recent trials in Tajikistan and Bolivia from McCarthy and colleagues). However we have done very minimal text modification here as these studies also had null findings for outcomes relevant to this study. In line with the comment from reviewer 2 below, we agree this reflects the general mixed to null findings from this space, which may be due to recruitment and implementation difficulties around implementing RCTs for digital health interventions (we are up front about our own challenges with contamination in the Discussion), but also that these may not be the most appropriate, stand-alone interventions for improving contraception-related outcomes among young people (an example of digital health enthusiasm having outpaced evidence).
4. I cannot comment on the statistical methods (someone more qualified could review) but good to state if the statistical analysis plan was pre-specified or not (sorry if I missed that) Thank you for noticing this omissionthe analysis plan was indeed pre-specified and we had failed to indicate this. We now do so in the last sentence of the 'datanalysis' subsection. 'Analyses were performed using Stata version 15, and all were conducted in accordance with a prespecified Statistical Analysis Plan.' Reviewer: 2 Dr. Julianne Weis, USAID Office of Population and Reproductive Health Comments to the Author: Good paper, contributes to growing evidence of mixed to null results of digital interventions in SRH knowledge of youth.
Our thanks for taking the time to review this paper and for the kind words.
Reviewer: 3 Dr. Narges Sheikhansari, University of Exeter Comments to the Author: Thank you for this very interesting study.
Our thanks for taking the time to review this paper. Reviewer: 1