The Effect of Mhealth Text Messaging Intervention on the Antenatal Care Contacts, Institutional Deliveries and Infants’ Birth-Weights of Expectant Adolescents and Young Mothers – A Quasi-Experimental Study in Kwale County, Kenya

Abstract

Background Poor referral channels and wanting knowledge on the significance of antenatal care have been shown to impede expectant adolescents and young women from having the recommended antenatal contacts. The ubiquity of mobile phones and the extensive use of text messages can improve ANC knowledge and hence increase antenatal contacts for expectant young women through prompt educational and reminder messages when integrated into existing antenatal service delivery points. Objective This study sought to determine the feasibility and acceptability of a mHealth-text messaging intervention. The specific objectives were to investigate the intervention's effect on antenatal care utilization, institutional deliveries, and infants' birthweights among young women aged 24 and below in Kwale County, Coastal Kenya. Study method This was a quasi-experimental trial using a mixed methods approach. The initial design was to have two arms, the mHealth text messaging intervention arm and the control arm. However, the low uptake of clinical services, a consequence of the COVID-19 pandemic forced some facilities in Kwale County to conduct clinical outreaches and to provide ANC and other maternal services at the community level. One control facility conducted monthly community outreaches which provided maternal related services to pregnant women. A substantial number (147 out of 213) of the control arm study participants in that facility were among those reached during outreaches. This situation was unexpected and had to be addressed by amending the control arm during data analysis. Thus, the original control group was split into “pure controls” and “outreach participants.” This resulted in having three study arms: (1) mHealth-text messaging intervention participants (2) pure control participants, and the (3) outreach arm participants. All quantitative data analyses were done using STATA version 15. Univariate, bivariate and multivariate regression analyses were done. Multivariate logistic regression with a 95% confidence interval was used to assess the relationships between independent variables and the study’s outcomes. Findings were presented as odds and risk ratios. Results were statistically significant at p<0.05. Additionally, a sensitivity analysis was done excluding Mkongani health centre from the analysis. Sensitivity analysis was done to estimate the consistency of the results using the ideal study subgroups, mHealth-text message vs pure control. Also, audio recordings from qualitative data were transcribed word for word and translated into English. This was followed by the identification of repeated patterns which were thematically analysed using NVivo version 12 Results A total of 817 participants were successfully linked to their background data and were included in the analysis. Findings from this study revealed that participants in the control arm had increased risk for low ANC contacts compared to those from the mHealth text-message intervention arm (Risk Ratio, RR: 0.5 [95% CI: 0.4 – 0.7], p<0.001) and outreach arm (RR: 0.1 [95% CI: 0.0 – 0.7] p = 0.024. It was also determined that, participants in the control arm had increased odds of unskilled deliveries compared to those in the mHealth-text message intervention arm (OR: 0.2 [95% CI: 0.1 – 0.4, p<0.001] and the outreach arm (OR: 0.1 [95% CI: 0.0 – 0.6], p=0.007). In contrast, babies born to participants from the mHealth-text message intervention (OR: 5.9 [95% CI: 1.7 – 20.8, p=0.006]) and outreach arms (OR: 6.6 [95% CI: 1.7 – 25.7], p=0.007) had significantly greater odds of low birth weight, compared to those born to control arm participants. On feasibility and acceptability of the mHealth intervention, this study found that mHealth text messaging intervention was feasible and acceptable to the young mothers; whereas interviewed healthcare providers were cautious with their views. Conclusion mHealth-text messaging is a feasible and acceptable intervention in the provision of ANC services targeting expectant young women. Together with community-based clinical outreaches, mHealth text messaging interventions can improve antenatal contacts and deliveries by a skilled healthcare practitioner. Policy makers and the public health community should consider implementing both mHealth and community clinical outreaches to improve maternal outcomes of young women in Kenya.