Abstract
BACKGROUND: Despite efforts to promote vaccination in low- and middle-income countries (LMICs), over 20 million infants remain under-immunized and at risk for unnecessary morbidity and mortality. Mobile health technologies, such as Short Message Service (SMS) texts, have tremendous and untapped potential for disease management. Patient reminder systems are an important mechanism for improving childhood vaccination coverage and can be easily adapted to SMS platforms. However, current research lacks an understanding of the barriers and facilitators to mHealth program design, implementation, and scale in LMICs. METHODS: We analyzed survey data collected March-November 2016 at the enrollment visit from a randomized controlled trial conducted at public health clinics in urban and rural Guatemala. Participants included eligible infants 6 weeks to 6 months of age receiving the first dose of the primary immunization series. At least one parent needed to own a mobile phone and be capable of deciphering SMS. Chi-square or Fisher's exact and Student's t-test were used to assess significance levels in demographic differences to describe factors that contribute to the feasibility of using an SMS-based vaccination reminder system. RESULTS: Of 1,088 families approached for enrollment, 871 were eligible and 720 (82.7%) participated with equal numbers of urban and rural children enrolled; 54 parents did not own a mobile phone with SMS capability and three parents could not use SMS. There was no significant difference between urban and rural maternal mobile phone ownership (94.4% vs. 93.3%, P=0.53), but more urban fathers owned mobile phones (72.8% vs. 47.1%, P<0.0001) and, overall, more mothers compared to fathers owned mobile phones (93.9% vs. 61.1%, P<0.0001). Most families (90.4%) chose to have reminders sent to the mother. Urban participants reported more mobile phones present in the home (P<0.0001), but rural participants reported more telephone landlines (34.7% vs. 15.6%, P<0.0001). Most participants reported a daily average of ≤5 telephone calls made (87.4%), ≤10 texts sent (91.0%), and ≤10 texts received (89.9%), with urban families reporting greater telephone usage (P=0.006, P<0.001, and P<0.001 respectively). Parents preferred to make calls over sending texts (74.7% vs. 25.3%, P<0.0001), with more urban families preferring text messaging (31.9% vs. 18.6%, P<0.0001). CONCLUSIONS: Our study results provide important insight into mobile phone access, usage, and preferences for voice and text communication across rural and urban populations of an LMIC that can be used to inform future mHealth interventions. Our findings suggest that offering a combination of more traditional communication methods with newer, modern technologies may be more effective at reminding families about vaccination visits, particularly for our rural population, and that targeting mothers for mobile phone interventions may provide the greatest benefits. Overall, our study suggests that using SMS reminders in LMICs can be a feasible tool for public health interventions.