Background
The Millennium Villages Project (MVP) implemented in Western Kenya a mobile Health tool that uses text messages to coordinate Community Health Worker (CHW) activities around antenatal care (ANC) and Prevention of Mother-to-Child Transmission of HIV (PMTCT), named the ANC/PMTCT Adherence System (APAS).

Methods
End-user changes in health-seeking behavior in ANC and postnatal care (PNC) were investigated following registration of 800 women into APAS. These investigations employed interviews of pregnant women or new mothers (n = 67) and CHWs (n = 20). Ordinal logistic regressions and exact binomial tests were used in the routine data analyses (n = 650, health registers).

Results
All CHWs interviewed agreed that APAS helped them track pregnant woman efficiently, compared to paper-based tracking forms. Women registered in APAS reported that CHWs reminded them of appointments more regularly than before its inception.

The routine data analysis showed that among women who had their 1st ANC visit in the 2nd trimester, women who resided in the MVP cluster and were in APAS had:
3 times the odds of going for more ANC visits compared to women who were not registered (but resided in the cluster), after adjusting for the mother’s HIV status in the multivariate model (Adjusted OR = 2.58, 95% CI [1.10-6.01]);
twice the odds of going for more ANC visits compared to women who were not registered and resided outside the cluster (Adjusted OR = 2.37, 95% CI [0.99-5.67])
Among women not registered, residence inside or outside the cluster did not affect the number of ANC visits made (Adjusted OR = 0.86, 95% CI [0.45-1.69]).

The APAS also greatly increased the likelihood of women making the 6 recommended post-delivery baby follow-ups.

For women registered in APAS, the MTCT rate at 18 months was significantly different from that of women not registered, and from the global rate of 30%. Women not registered had a 9% MTCT rate at 18 months regardless of residence, while women registered had a 0% transmission rate at both 9 and 18 months.

Conclusions
The incorporation of mHealth tools in CHW programs can improve adherence to ANC and PNC and enhance PMTCT efforts.