Information and communication technologies (ICT), including mobile phones (eg, smartphones and regular mobile phones), Internet, email, and social networking have the potential to improve care for underserved communities with reduced access to health care. Although pregnant women frequently utilize health care services, they are often lost to follow-up after delivery, even among those with high-risk pregnancies [1]. Prior studies in pregnant and postpartum women show increasing use of Internet for delivery of health information [2] and interest in mobile phone apps [3]. Potential participants in postpartum weight loss interventions report high interest in Web-based components [4]. ICT has the potential to engage with and educate women, thus promoting improved health for women and their families before and after delivery [5], but evidence to support the uptake among low-income racially and ethnically diverse populations is limited.

The Pew Research Center showed 90% of Americans used a mobile phone [6] and 87% reported at least some Internet use [7]. Among those who report Internet use, 91% send or read emails and 74% use social networking sites such as Facebook, Myspace, or LinkedIn. Those with higher education and in younger age groups report most frequent Internet use, suggesting that ICT may be an optimal tool for communication, monitoring, education, and even providing interventions to young women [7]. However, studies have shown that low-income populations have fewer available ICT resources, including a reliable home Internet connection, and also frequently change mobile phone numbers [8,9]. People with limited English proficiency face multiple barriers to health care and community services, indicating a high need for novel and effective outreach strategies to engage and communicate with this population [10-14].

To inform the development of ICT-based health-related programming to a diverse population of pregnant and postpartum women, we conducted a cross-sectional survey about their ICT usage. The aim of this study was to evaluate differences in the use of ICT between racial and ethnic groups and by English language proficiency. We hypothesized that pregnant and postpartum women frequently use ICT, including mobile phones, Internet, and social networking, and that rates of ICT use would be lowest among Latinas and those with limited English language proficiency. The rationale for the hypothesis was based on prior literature indicating that Latino adults, especially those with less education, had lower Internet usage [15].

We designed the study to inform the development of culturally appropriate ICT interventions to promote healthy lifestyle behaviors in the perinatal period.

In this sample of low-income, racially and ethnically diverse pregnant and postpartum women, mobile phone and SMS text message usage were common across all racial/ethnic groups. Although more than 85% of all participants reported SMS text messaging, African American women were less likely to text compared to white women. The rates of mobile phone usage reported in our survey were similar to The Pew Research Center’s Internet Project Survey data showing 90% of people own cell phones, including 90% of whites and African Americans and 92% of Hispanics [6]. Additionally, we confirmed low rates of landline use (57% did not own landlines) consistent with Pew Internet Research findings (41%). In the Pew survey, households with lower socioeconomic status and Hispanics had higher rates of not having landlines (56.2% and 53.1%, respectively) [20]. Although our sample generally reported high mobile phone and SMS text messaging rates, we identified disparities in Internet, email, and social networking use by racial/ethnic groups and limited English language proficiency. Our results suggest that mobile phones are potentially useful modalities for the delivery of health interventions to low-income pregnant and postpartum women, but interventions requiring Web-based apps may have lower uptake unless alternatives (ie, paper) and Spanish translations are available.

Other studies have also reported lower rates of Internet usage among Latinas, but without close examination of the role of English language proficiency. A large cross-sectional survey of 3181 young women attending reproductive health clinics in Texas reported 92.7% of whites and 92.9% of African American women, but only 67.5% of Latinos, used the Internet. Hispanic women reported barriers to Internet use including cost, not having a computer at home, and not knowing how to use a computer [21]. However, the impact of English language proficiency on differential ICT use was not described. Notably, more than 80% of Latinas in our study had limited English language proficiency and 17% spoke any English, which is in contrast to the Latino population sampled in the Pew Internet Research Survey in which 65% were English speaking (either English dominant or bilingual) and proficiency was not assessed. The large difference in English proficiency likely accounts for the disparity we noted in Internet use, in which only 62% of Latinas in our sample used the Internet compared to 78% of Latinos in the Pew Survey [22]. In addition, more than half of the Latinas in our sample reported less than a high school education; from other surveys from this community, more than one-third likely have less than a sixth grade education [15] indicating lower literacy, including Spanish literacy. Our results suggest that English language and literacy are major barriers for women to use and access the Internet. The development of technology-based interventions, especially those that require Internet components, should be translated into Spanish, designed for people with lower literacy, and culturally adapted for Latinos to have the greatest potential impact.

Despite these disparities in Internet usage, our study supports mobile phone-based interventions in a low-income, racially diverse population. Growing evidence supports mobile phone-based interventions to impact health behaviors, but few studies have focused on pregnancy and postpartum health [23]. One example of a large-scale SMS text messaging program aimed at improving prenatal care and pregnancy outcomes is the Text4Baby program launched in 2010 by the Centers for Disease Control and Prevention. Women who sign up receive texts containing information about prenatal and postpartum health behaviors and services [24]. However, a randomized controlled trial of 123 women (approximately 80% of whom were Spanish speaking and 75.6% who had participated in the Special Supplemental Nutrition Program for Women, Infants, and Children [WIC] program) did not show a difference between text4baby intervention and usual care control in terms of changes in self-reported health behaviors, but no birth or utilization of care outcomes were reported [25]. Adaptation of the SMS text messaging programs for low-income women is especially important. The WIC program serves low-income women and children and has also been focused on improving care delivery through mobile-based apps [25,26]. A WIC program in Atlanta, GA, tested the text4baby program to assess participants’ enrollment and satisfaction in 468 (91% African American) participants. Only 51% of women provided with enrollment instructions attempted to enroll in the program; among these, 69% successfully enrolled mostly via SMS text message (vs online). Higher education and higher incomes were associated with increased enrollment, indicating that the enrollment process may have more barriers for less educated and poorer women [27]. This study notes the importance of testing actual use of ICT interventions in low-income populations to reduce their risk of widening the disparities that they were designed to address.

An additional challenge to implementing and sustaining SMS text messaging interventions and programs is the frequency with which women reported changing mobile phones or phone numbers. In our sample, more than one-quarter of women reported having 2 or more cell phone numbers in the last 12 months, with the highest rates among African American women (35%) and Latinas (25%). To facilitate intervention adherence involving use of mobile phones, studies have budgeted funding to provide mobile phone minutes or plans or even phones to participate to enhance participation rates, but this may not be cost-effective for community-based programs and the aforementioned intervention was not offered in multiple languages [28]. Other studies have required an unlimited short messaging plan [29], but this may exclude lower income populations [30].

The major strengths of our study are including both obstetric and pediatric sites, and identifying a higher risk population of women with medical comorbidities and a racially and ethnically diverse sample. We also collected information about perceived English language proficiency to analyze the results not only by racial and ethnic groups, but also by English language proficiency.

There are several limitations of our study. First, we surveyed a convenience sample of women who were attending one of several clinical sites and may have missed women who did not receive prenatal care or who did not attend visits. This may have made the study’s results less generalizable to other women in Baltimore, but because we collected data over several months, women had multiple opportunities to attend visits and complete the surveys. Second, because this was a self-administered questionnaire, we screened out 5 participants who had self-reported low literacy in English or Spanish as part of study eligibility and thus our results provide a “best-case scenario” for Internet and mobile phone use among literate women. Third, the cross-sectional design limits our ability to assess causality. Fourth, our study examined racial and ethnic differences in Internet and mobile phone usage in a population of low-income women in Baltimore, MD, and these differences may be different in other cities in the United States. Fifth, there is a potential that results being attributed to ethnicity may in fact be due to income instead, especially because the latter was not included in the multivariate regression analysis. Sixth, our study did not account for potential differences in ICT usage for nulliparous versus multiparous women because more experienced mothers may differ in their level of need for information.

In conclusion, our findings show that the racial and ethnic digital divide regarding mobile phone use and SMS text messaging is diminishing, but persists for Internet, email, and social networking by race and ethnicity and particularly for women with limited English language proficiency. These findings support developing linguistically and culturally appropriate mobile phone and SMS text messaging interventions for women of all ethnicities and language proficiencies to promote improved healthy lifestyle behaviors, specifically in the perinatal period.