Early accurate diagnosis is one of the first crucial steps in care for infants born to HIV-infected mothers. Early initiation of antiretroviral therapy (ART) relies upon early diagnosis and results in significant reductions in infant morbidity and mortality. The investigators recently concluded a successful randomized controlled trial in Kenya entitled, "Improving uptake of early infant diagnosis of HIV for prevention of mother-to-child HIV transmission (PMTCT): a randomized trial of a text messaging intervention" (ClinicalTrials.gov # NCT01433185). In this study, text messages developed using a behavioral theoretical framework significantly improved maternal attendance at post-partum clinic appointments and rates of testing to facilitate early infant diagnosis of HIV in a selected population and controlled setting. Understanding the effectiveness of this intervention (and its limitations) in a real-world, routine-care setting represents the next step in the translational pathway to public health impact. The investigators therefore now propose a cluster randomized, stepped wedge trial in 20 clinics operated by the Kenyan Ministry of Health in the Nyanza region of Kenya and use the Reach, Effectiveness, Adoption, Implementation, and Maintenance (REAIM) framework to understand the effectiveness of the text messaging to improve testing (TextIT) intervention. Our specific aims are:

1. To determine the effect of TextIT on maternal attendance at postpartum clinic visits during the randomized stepped-wedge rollout of the intervention.

   Hypothesis 1: A greater proportion of women at health facilities implementing TextIT will attend clinic within eight weeks postpartum compared to women at health facilities implementing standard care.

2. To determine the effect of TextIT on virological infant HIV testing within eight weeks after birth during the randomized stepped-wedge rollout of the intervention.

   Hypothesis 2: Infants of women at health facilities implementing TextIT will be more likely to have virological HIV testing compared to infants of women at health facilities implementing standard care.

3. To determine the costs and cost-effectiveness of TextIT. The investigators will estimate the cost per patient and per health gain achieved (disability-adjusted life year, DALY) comparing TextIT to current standard care.

Hypothesis 3: The TextIT intervention will be more cost-effective than current standard care.

The 2011 United Nations General Assembly Political Declaration on HIV/AIDS set a target to eliminate mother-to-child transmission of HIV by 2015. The continuum of care for women living with HIV and their babies includes retention in care and early infant diagnosis of HIV. Timely initiation of infant antiretroviral therapy (ART) requires HIV-positive women to be retained in care through the postpartum period and to bring their children for HIV testing. However, a high proportion of pregnant HIV-positive women in sub-Saharan Africa are lost to follow-up after delivery. The World Health Organization recommends infant HIV testing at six weeks using DNA polymerase chain reaction (PCR). However, the overall proportion of children who undergo early infant diagnosis by PCR remains low. In Kenya, the overall proportion of eligible children undergoing PCR testing in 2011 was 39%.

Low rates of infant HIV testing are an indirect indication of a large number of infants who may not benefit from early determination of HIV status, antiretroviral prophylaxis for HIV-negative breastfeeding infants, and infant feeding counseling and support. For HIV positive infants, failure to undergo testing is a critical barrier to receiving life-saving ART. There is an urgent need to investigate efficacious, cost-effective, and sustainable interventions to improve maternal retention in HIV care and increase the proportion of exposed infants tested for HIV.

With the exponential increase in the number of mobile phones in sub-Saharan Africa, the use of mobile technology to support HIV programs has shown promise. The United Nations Children's Fund is encouraging HIV programs to take advantage of high levels of mobile phone access among enrolled mothers by reminding them to return for critical appointments. Active follow-up of pregnant HIV positive women in resource-limited settings using mobile phones is feasible, and may be effective in improving postnatal retention and uptake of early infant diagnosis across different sub-Saharan African settings.

Inclusion Criteria:

• Are ≥18 years or emancipated minors;
• Are at 28 weeks gestation or greater (or have delivered on the day of enrollment);
• Provide informed consent

Exclusion Criteria:

• Women who report sharing phones but have not disclosed their HIV status to the person with whom the phone is shared.