The readiness, need for, and effect of mhealth interventions to improve immunization timeliness and coverage in rural western Kenya
Embargo until
Date
2014-12-23
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Johns Hopkins University
Abstract
Background: As mobile phone ownership levels grow globally, opportunities to target hard-to-reach populations through mobile-health technologies become more realistic. In a rural Kenyan population, this dissertation seeks to assess the magnitude and risk factors of immunization timeliness and coverage, to determine the readiness of a community for mHealth interventions by assessing prevalence and risk factors for mobile phone ownership and SMS utilization, and to assess the effect of the Mobile Solutions for Immunization (M-SIMU) cluster randomized controlled trial.
Methods: A cross-sectional survey of Kenyan caregivers was conducted to collect baseline immunization and mobile phone ownership estimates for M-SIMU. Predictors of mobile phone ownership were obtained through multivariable logistic regression. Risk factors for delayed immunizations and not receiving immunization were calculated using binomial regression with log link. The M-SIMU trial randomized villages to four arms: Control, short message system (SMS) reminders only, SMS reminders + 75 Kenyan Schillings (KSH) incentive or, SMS reminders + 200 KSH incentive. Inverse Kaplan-Meier curves and Cox regressions assessed the intervention’s effect on pentavalent3 and measles vaccination.
Results: Older maternal age, higher maternal literacy and education, smaller households, and higher socioeconomic status were associated with phone ownership. Immunization coverage for the third dose of pentavalent vaccine (pentavalent3), measles, and fully immunized children (FIC) were 95%, 83%, and 80%, respectively. Delayed pentavalent1 was associated with not receiving pentavalent3 (RR: 5.61; 95%CI: 3.77-8.33), measles vaccine (RR: 1.51; 95%CI: 1.15-1.99), and FIC (RR: 1.87; 95%CI: 1.51-2.32). The prevalence of delayed pentavalent1, pentavalent3, and measles were 11%, 24%, and 29%, respectively. No common risk factors in the delay models were found. For M-SIMU, Kaplan-Meier curves found significant differences across arms in time to pentavalent3 (p<0.01) but not measles vaccination (p=0.10). SMS + 200 KSH infants were associated with pentavalent3 vaccination (HR: 3.33; 95%CI: 1.71-6.47) and approached significance for measles (HR: 2.05; 95%CI: 0.95-4.41; p=0.07), as compared to controls. The SMS only and SMS plus 75KSH were not significantly associated with either vaccine.
Conclusions: In a population with moderate phone ownership, high immunization coverage, and moderate vaccine delays, SMS reminders plus 200KSH improved pentavalent3 vaccination.
Description
Keywords
vaccines, mHealth