i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 4 ( 2 0 1 5 ) 341–348
journal homepage: www.ijmijournal.com
Disparities in mobile phone access and maternal health service utilization in Nigeria: A population-based survey Larissa Jennings a,b,∗ , Adetayo Omoni b,c , Akunle Akerele c , Yisa Ibrahim c , Ekpenyong Ekanem b,c a
Johns Hopkins Bloomberg School of Public Health, Department of International Health, 615N. Wolfe Street, Room E5038, Baltimore, MD 21205, USA b Abt Associates, Inc., International Health Division, Monitoring and Evaluation, 4550 Montgomery Ave, Suite 800N, Bethesda, MD 20814, USA c Partnership for Transforming Health Systems II (PATH2), Monitoring and Evaluation, 37 Panama Street, IBB Way, Maitama, Abuja, Nigeria
a r t i c l e
i n f o
a b s t r a c t
Article history:
Background: Mobile communication technologies may reduce maternal health disparities
Received 11 July 2014
related to cost, distance, and infrastructure. However, the ability of mHealth initiatives to
Received in revised form
accelerate maternal health goals requires in part that women with the greatest health needs
14 January 2015
have access to mobile phones.
Accepted 22 January 2015
Objective: This study examined if women with limited mobile phone access have differential odds of maternal knowledge and health service utilization as compared to female mobile
Keywords:
phone users who are currently eligible to participate in maternal mHealth programs.
Mobile phones
Methods: Using household survey data from Nigeria, multivariable logistic regressions were
Maternal health
used to examine the odds of maternal knowledge and service utilization by mobile phone
Service utilization
strata.
Access
Results: Findings showed that in settings with unequal access to mobile phones, mHealth
Equity
interventions may not reach women who have the poorest maternal knowledge and care-
Disparities
seeking as these women often lacked mobile connectivity. As compared to mobile users, women without mobile phone access had significantly lower odds of antenatal care utilization (OR = 0.48, 95%CI: 0.36–0.64), skilled delivery (OR = 0.56, 95%CI: 0.45–0.70), and modern contraceptive use (OR = 0.50, 95%CI: 0.33–0.76) after adjusting for demographic characteristics. They also had significantly lower knowledge of maternal danger signs (OR = 0.69, 95%CI: 0.53–0.90) and knowledge of antenatal (OR = 0.46, 95%CI: 0.36–0.59) and skilled delivery care benefits (OR = 0.62, 95%CI: 0.47–0.82). No differences were observed by mobile phone strata in uptake of emergency obstetric care, postnatal services, or breastfeeding. Conclusions: As maternal mHealth strategies are increasingly utilized, more efforts are needed to improve women’s access to mobile phones and minimize potential health inequities brought on by health systems and technological barriers in access to care. © 2015 Elsevier Ireland Ltd. All rights reserved.
∗
Corresponding author at: Johns Hopkins Bloomberg School of Public Health, Department of International Health, 615 N. Wolfe Street, E5038, Baltimore, MD 21205, USA. Tel.: +1 410 955 3537. E-mail addresses: [email protected], [email protected] (L. Jennings), adetayo [email protected] (A. Omoni), [email protected] (A. Akerele), [email protected] (Y. Ibrahim), ekpenyong [email protected] (E. Ekanem). http://dx.doi.org/10.1016/j.ijmedinf.2015.01.016 1386-5056/© 2015 Elsevier Ireland Ltd. All rights reserved.
342
1.
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 4 ( 2 0 1 5 ) 341–348
Introduction
Mobile communication technologies have the potential to reduce maternal health disparities related to cost, distance, and inadequate health infrastructure, even in technologypoor settings [1–3]. Yet, the ability of mobile health (mHealth) initiatives to accelerate maternal development goals requires in part that women with the greatest health needs have access to mobile communication devices [4,5]. While mobile phone coverage in developing countries has rapidly increased in recent years [3,6], research has shown that significant disparities in access remain by gender, household, and geographic characteristics [4,7,8]. Women account for almost two-thirds of the global mobile phone market that has not yet been reached [7], and those who are the poorest and least educated are most often lacking connectivity [3,4,7]. Poor and uneducated women also have higher rates of prenatal, intranatal, and postnatal morbidity and mortality [9–11]. While substantial gains have been made in addressing maternal morbidity and mortality, some achievements have resulted historically in greater health inequity at the population level with significantly lower uptake among the most vulnerable and marginalized women [12–15]. This has raised concerns whether the expansion of maternal mHealth applications that directly target patients-such as public health communications, personalized behavior change messages, or self-care information is likely to exacerbate health disparities among women most at-risk for adverse health outcomes [2,4,16]. Implementing equitable maternal mHealth interventions requires information on women’s differential health needs by mobile phone strata [3–5,13,17,18]. However, few studies have examined need-related inequalities in maternal service utilization as compared to mobile phone access. This has limited the ability of mobile-based maternal health programs to assess beforehand the potential of serving women and families most at-risk. Given that mHealth applications are increasingly being utilized to improve the delivery of maternal health services, this analysis aimed to preliminarily inform the development of a community-based mHealth initiative targeting women with low maternal knowledge and care-seeking behaviors. We sought to determine whether a mHealth intervention aiming to increase uptake of maternal services through participant interaction was likely to reach women with the highest unmet maternal need (i.e., low antenatal use, unskilled delivery attendance, poor maternal knowledge, etc.), as indicated by access to a mobile phone. Specifically, we examined if women with limited mobile phone access had differential odds of maternal health service utilization or knowledge of pregnancy, birth, and postpartum health as compared to female mobile phone users who are currently eligible to participate in maternal mHealth programs. Implications on equitable implementation of mobile technology to reach the world’s most vulnerable women are discussed.
2.
Methods
2.1.
Sampling
Data were used from the Partnership for Transforming Health Systems, Phase Two (PATHS2) project’s baseline household
survey in Nigeria. Using a population-based, two-stage cluster design, the PATHS2 household survey was conducted in five project states: Enugu, Jigawa, Kaduna, Kano, and Lagos to measure pre-intervention prevalence of a range of maternal, reproductive, and child health outcomes. In the first stage, 625 enumeration areas were selected from the local government areas within the five states. In the second stage, a target sample of 10,000 households was selected from the enumeration areas. Local and trained interviewers then surveyed the head of household and, if applicable, one woman aged 15–49 with a child 23 months or less was randomly selected from the household.
2.2.
Maternal health setting
Despite declines in recent years, Nigeria accounts for an estimated 14% of maternal deaths worldwide, the second largest contributor to maternal mortality in the world [19]. In 2013, an estimated 40,000 Nigerian women died in pregnancy and birth-related complications representing approximately 560 maternal deaths for every 100,000 live births [20]. Care from a skilled provider during pregnancy and childbirth can reduce the risk of complications and infections that lead to illness or death of mothers and newborns [21–23]. Yet, only 61% of women in Nigeria reported receiving antenatal care from a trained provider during their most recent pregnancy, and only 38% of Nigerian births were delivered by a skilled health provider [22]. As of 2014, there were an estimated 129 million mobile phones in use in Nigeria, representing about 76% of the total population [24]. However, mobile phone coverage varies substantially across the country, and women are significantly less likely than men to own a mobile phone [25–27].
2.3.
Data collection
Data were collected in July and August 2012. A structured questionnaire was used to elicit information on household demographics and women’s maternal and reproductive histories. The questionnaire was developed in English and blind, back-translated into Yoruba, Igbo, and Hausa, the languages spoken predominantly in the study areas. All survey questions were administered orally using personal digital assistants in a language chosen by the participant. Interviewers were trained on the survey’s objective and methodology using didactic methods as well as mock interviews and pilot tests.
2.4.
Outcome measures
The outcome measures for the study comprised maternal knowledge and health service utilization during three periods: pregnancy, labor and delivery, and the early postnatal period. Two care-seeking behaviors were assessed during pregnancy. Women were asked if they received antenatal care during their last pregnancy: yes (code = 1) or no (code = 0). For women who received antenatal care, we also dichotomized the sample among those who received the minimum recommended four antenatal care visits (code = 1) versus those who received three or fewer visits (code = 0). To assess maternal knowledge regarding care during pregnancy, women were asked to describe the advantages of attending antenatal care.
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 4 ( 2 0 1 5 ) 341–348
Responses were matched against five pre-coded antenatal care benefits: receipt of health education, examination by a skilled provider, early detection and management of complications, receipt of drugs/medicines, malaria prevention, and anemia prevention. An antenatal benefit corresponding to one of the five responses generated a score of 1, where a knowledge score ≥ 3 was coded as 1 compared to a knowledge score < 3 (code = 0). Women were also asked what danger signs would indicate that a woman who is pregnant, in labor, or has recently delivered needs to be rushed to a health facility for emergency care. Women who knew of at least three maternal danger signs (severe headache, swelling, dizziness, fever/chills, convulsions, hemorrhage, prolonged labor, malpresentation, or retained placenta) were coded as 1 compared to women who were less knowledgeable, coded as 0. Four health care-seeking scenarios were examined pertaining to labor and delivery. Women were asked who attended the delivery of their last child: a doctor, nurse/midwife, or auxiliary midwife (code = 1) versus those with unskilled assistance (code = 0). Women reported likewise if they experienced any complication during their last child’s birth (severe headache, convulsions, hemorrhage, prolonged labor, mal-presentation, or retained placenta): yes (code = 1) or no (code = 0). If during labor and delivery of their last child an ambulatory transportation service was contacted for emergency travel to a health center, women were coded as yes (code = 1) or no (code = 0). To examine maternal knowledge regarding delivery care, women were asked to give two reasons why delivering in a health facility was beneficial: ≥ 2 responses out of the nine precoded responses (receipt of health education, management of obstetric complications, prevention of infection, prevention of vertical HIV transmission, immediate newborn care, provision of other necessary treatment/drugs, avoidance of emergency care delays, health provider assistance, and psychosocial support) were coded as 1, compared to < 2 responses, coded as 0. Two questions on postnatal care utilization examined whether a doctor, nurse/midwife, or health center staff visited the woman at home in the first week after her most recent birth: yes (code = 1) or no (code = 0). If yes, the woman was also asked how many days after birth did a skilled provider first visit her home: within a day (code = 1) or within 2 to 7 days (code = 0). Women were also asked if they ever breastfed their youngest child, and if yes, how long after birth did they initiate breastfeeding: within the first hour of birth (code = 1) versus > 1 hour (code = 0). Lastly, all women indicated whether they were currently using any method to delay or avoid pregnancy: modern contraceptive method (code = 1) versus those who were not (code = 0). Information was also obtained on each woman’s age, highest level of education attained, location (rural/urban), state, and household wealth index.
2.5.
Exposure measure
The exposure variable of interest was mobile phone status. Women were asked if they currently had access to a mobile phone, defined as having the option to use in a singular or shared fashion a functional cellular phone. Women who responded “yes” were categorized as the reference group
343
(code = 0) compared to women who did not have mobile phone access (code = 1).
2.6.
Data analysis
Data were analyzed using STATA, Version 12 (Stata Corporation, College Station, TX). Analyses were done with sampling weights using the svy command in STATA to account for the survey cluster design. We first examined the distribution of sample demographic characteristics and outcomes by mobile phone status using two-sample tests of proportions and chisquare (2 ) statistics. Next, bivariate and multivariable logistic regression models were used to examine differences in the odds of maternal knowledge and service utilization by mobile phone status. Odds ratios were considered statistically significant at p < .05, when the null value of 1.0 was not included in the corresponding 95% confidence interval.
3.
Results
3.1.
Sample characteristics
A total of 3,390 women aged 15–49 with a child 23 months or less were identified by the survey (Table 1) out of a final sample of 10,107 households. The mean age of women was 28.3 years, and 51.3% had received primary education or higher. Nearly two-thirds (61.3%) of women lived in a rural area, and those from Kano (33.8%) and Lagos (24.1%) states made up the largest proportions of the total sample. Mobile phone access varied by demographic characteristics. Approximately half (51.4%) of women did not have access to a mobile phone. As compared to those who reported having mobile phone access, women who did not were significantly younger (27.2 years vs. 29.4 years, p < 0.001), had lower rates of primary educational attainment (27.2% vs. 66.1%, p < 0.001), and were more likely to live in a rural setting (82.5% vs. 37.0%, p < 0.001). Women without access to a mobile phone also made up greater proportions of the poorer and poorest household wealth quintiles (72.0% vs. 16.5%, p < 0.001).
3.2.
Antenatal care
Seventy-five percent (75.0%) of women received antenatal care by a skilled provider at least once during their most recent pregnancy, and 69.1% received all four recommended antenatal visits (Table 2). Antenatal care utilization varied considerably by mobile phone status. In bivariate models, women without access to a mobile phone had 83% significantly lower odds (OR = 0.17, 95%CI: 0.14–0.22) of receiving any skilled antenatal care, and 53% significantly lower odds (OR = 0.47, 95%CI: 0.38–0.58) of having at least four skilled antenatal care visits (Table 3). These differences remained after controlling for demographic characteristics. Compared to mobile phone users, the adjusted odds of attending one or all of the recommended four antenatal care visits was 52% lower (OR = 0.48, 95%CI: 0.36–0.64) and 29% lower (OR = 0.71, 95%CI: 0.55–0.92), respectively, among women without access to a mobile phone.
344
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 4 ( 2 0 1 5 ) 341–348
Table 1 – Demographic characteristics for total sample and by mobile phone access. By Mobile Phone Access
Total
Frequency (n = ) Proportion of sampled womena (%) Maternal age in years (mean ± SE) Education Primary or more Less than primary Location Rural Urban State Enugub Jigawa Kaduna Kano Lagos Household wealth quintile Poorest Poorer Middle Richer Richest a b ∗
3930 100.0 28.3 (±0.1)
With
Without
p-Value
1811 46.8 29.4 (±0.2)
1998 51.4 27.2 (±0.2)
– – 0.000*
51.3 43.5
76.9 19.7
27.2 66.1
0.000*
61.3 38.7
37.0 63.0
82.5 17.5
0.000*
5.9 21.9 14.4 33.8 24.1
8.2 13.6 11.6 20.3 46.4
0.9 30.2 17.3 47.1 4.6
0.000*
23.8 22.5 17.6 18.4 17.6
6.4 11.1 17.5 29.4 35.6
39.9 32.9 17.1 8.6 1.6
0.000*
Population estimates are weighted to adjust for cluster survey design. Phone access data missing for 26.8% of sample. Significant at p < 0.05.
Twenty-seven percent (27.2%) of women knew three or more benefits to antenatal care (Table 2). Even fewer (18.9%) were knowledgeable of at least three maternal danger signs requiring urgent care. After controlling for demographic characteristics, women without mobile phone access had 54% significantly lower odds (OR = 0.46, 95%CI: 0.36–0.59) than female mobile phone users of knowing the benefits of antenatal care (Table 3). In multivariable analyses, women lacking mobile connectivity also had 31% significantly decreased odds
(OR = 0.69, 95%CI: 0.53–0.90) of correctly identifying pregnancy danger signs compared to mobile-connected women.
3.3.
Labor and delivery
Similar patterns were observed in labor and delivery care utilization by mobile phone status. While nearly half (48.3%) of women delivered with a skilled health provider (Table 2), women who lacked access to a mobile device had an 82%
Table 2 – Maternal service utilization and knowledge for total sample and by mobile phone access. Proportion of sampled women (%)
Frequency (N=) Antenatal care Received at least 1 antenatal care visit Received recommended 4 antenatal care visits Maternal knowledge of benefits of antenatal care Maternal knowledge of obstetric danger signs Labor and delivery Delivered with skilled birth attendant Had obstetric complication Used ambulatory transportation for obstetric emergency Maternal knowledge of benefits of skilled birth attendance Postnatal care Received at least one postnatal visit within one week Received at least one postnatal visit within day of birth Initiation of breastfeeding within one hour of birth Uptake of modern contraception a ∗
Population estimates are weighted to adjust for cluster survey design. Significant at p < .05
By Mobile Phone Access
Total With
Without
1811
1998
75.0 69.1 27.2 18.9
90.1 76.2 38.7 22.8
61.2 60.0 16.7 15.3
0.000* 0.000* 0.000* 0.000*
48.3 10.3 3.7 74.9
69.2 11.3 5.1 86.1
28.7 9.5 2.3 64.8
0.000* 0.141 0.000* 0.000*
10.1 32.7 40.8 7.5
12.3 31.9 37.9 12.4
8.1 33.9 43.6 3.1
0.002* 0.759 0.012a,* 0.000*
3930
p-Value –
345
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 4 ( 2 0 1 5 ) 341–348
Table 3 – Odds ratios and 95% confidence intervals of maternal service utilization for women without mobile phone access compared to reference group. Bivariate modela,c
Total: N = 3930 OR Antenatal care Received at least 1 antenatal care visit Received recommended 4 antenatal care visits Maternal knowledge of benefits of antenatal care Maternal knowledge of obstetric danger signs Labor and delivery Delivered with skilled birth attendant Had obstetric complication Used ambulatory transportation for obstetric emergency Maternal knowledge of benefits of skilled birth attendance Postnatal care Received at least 1 PNC visit within one week Received at least 1 PNC visit within day of birth Initiation of breastfeeding with one hour of birth Uptake of modern contraception a b
c ∗
95% CI
Multivariable modelb,c OR
95% CI
0.17* 0.47* 0.32* 0.61*
0.14–0.22 0.38–0.58 0.26–0.39 0.49–0.75
0.48* 0.71* 0.46* 0.69*
0.36–0.64 0.55–0.92 0.36–0.59 0.53–0.90
0.18* 0.82 0.44* 0.30*
0.15–0.22 0.64–1.07 0.29–0.68 0.24–0.37
0.56* 0.80 0.82 0.62*
0.45–0.70 0.58–1.11 0.44–1.53 0.47–0.82
0.63* 1.09 1.27* 0.23*
0.48–0.84 0.62–1.92 1.06–1.53 0.16–0.32
1.21 0.72 1.02 0.50*
0.83–1.76 0.35–1.48 0.81–1.27 0.33–0.76
Bivariate model adjusted for cluster survey design only. Multivariable model adjusted for cluster survey design and control variables [maternal age, education, location (rural/urban), state, and household wealth] with mean variance inflation factors ranging: 1.27 to 1.85. Index group = women without mobile phone access (code = 1); Reference group = women with mobile phone access (code = 0). Significant at p < .05.
significant decrease (OR = 0.18, 95%CI: 0.15–0.22) in the odds of delivering with a skilled attendant as compared to those with mobile phone access (Table 3). This relationship was maintained at 44% significantly lower odds (OR = 0.56, 95%CI: 0.45–0.70) of delivering with a skilled provider after controlling for demographic factors. Most surveyed women (74.9%) were knowledgeable of at least two benefits to having skilled delivery care (Table 2), but women without mobile phone access were 38% significantly less likely (OR = 0.62, 95%CI: 0.47–0.82) to be aware as compared to their mobile counterparts in multivariable models (Table 3). In contrast, no significant differences were observed in multivariable models in the proportion of women who had a complication during childbirth by mobile phone status (OR = 0.80, 95%CI: 0.58–1.11), nor in ambulatory transportation use for obstetric emergencies (OR = 0.82, 95%CI: 0.44–1.53).
3.4.
Postnatal care
Postnatal care utilization was lower among surveyed women than antenatal care utilization. A third (32.7%) of women received at least one postnatal visit at home within one day of birth, and 10.1% within the first week after delivery (Table 2). No significant differences were observed in multivariable models in the odds of postnatal care utilization by phone status within a day of birth (OR = 0.72, 95%CI: 0.35–1.48) or within one week following birth (OR = 1.21, 95%CI: 0.83–1.76) (Table 3). Among postnatal care practices, 40.8% of women reported initiating breastfeeding within the first hour of birth, and women lacking mobile connectivity had a 1.27 times significantly greater odds of immediate breastfeeding (OR = 1.27 95%CI: 1.06–1.53). However, this association was not statistically significant in multivariable analyses (OR = 1.02, 95%CI: 0.81–1.27). Uptake of modern contraceptive methods was low (7.5%) among all women, but women who did not
have access to a mobile phone had a significant 50% decrease (OR = 0.50, 95%CI: 0.33–0.76) in the odds of using any modern contraceptive method, after controlling for demographic characteristics.
4.
Discussion
The ability of mHealth initiatives to accelerate maternal development goals requires in part that women with the greatest health care needs have access to mobile devices. Yet, few studies have examined maternal health priorities by mobile phone status. This study found that after controlling for demographic factors, women with no mobile phone access were significantly less likely to utilize skilled antenatal and delivery care services as compared to mobile-connected women who are currently eligible to participate in patient-based maternal mHealth interventions. Women without access to mobile phones were also significantly less knowledgeable of the benefits of skilled maternal care and dangers signs requiring urgent medical attention. Such findings suggest that in settings with unequal access to mobile phone technologies, mHealth programs that aim to increase uptake of maternal services through direct support to patients may be less effective in reaching women with the poorest maternal knowledge and care-seeking as these women often lacked access to mobile phones. Other studies have also highlighted this weakness [28]. A recent review of mHealth interventions in sub-Saharan Africa found that mHealth initiatives aiming to disseminate health education had been less successful due to limited accessibility among the most vulnerable groups [29]. In Uganda, evaluators of an HIV/AIDS text-messaging campaign indicated also that the intervention failed to provide information to the most vulnerable women, those with the lowest income and who did not own mobile phones [30].
346
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 4 ( 2 0 1 5 ) 341–348
Given the growing need to ensure that the benefits of maternal mHealth strategies are equitably distributed, this paper offers several considerations for enhancing future practice and design. One important strategy for mHealth implementers who are targeting the most vulnerable women is to assess during the development phase the potential distributional impact of mobile-based strategies on maternal outcomes. Careful planning is needed in determining whether the intended mobile technology is accessible to women with the greatest maternal health needs. While the process of measuring equitable impact is complex [14], the results from this study suggest that a simple, pre-intervention assessment can provide important insights during programmatic design on differences in maternal health needs by mobile phone status. Such information may be used by implementers to tailor maternal mHealth programs with technologicallyappropriate components for women with varying levels of mobile connectivity. A second implication is to improve women’s access to mobile phones, since women who lacked mobile access in this study represented the population most likely to benefit from improved maternal health interventions. Research has shown that expanding access to mobile communication devices among women in resource-poor settings can yield several social, economic, and health benefits [7]. This may reflect empowerment differences in women who were able to access and utilize new technologies. On the other hand, empowering women by improving access to mobile technologies may strengthen their capacity to utilize maternal health-related services, including skilled antenatal and delivery care, and increase their participation in social and economic activities that positively impact reproductive decision-making. Results from this study suggest that the ubiquity of mobile phones is not universal. Greater numbers of women will likely need access to mobile phones in order to improve the capacity of mHealth strategies to minimize rather than exacerbate maternal health disparities in resource-poor settings. In contrast, given that roughly half of the surveyed women had access to mobile phones, this study also highlights the possibility of engaging current female mobile phone users as transfer agents of mobile-based outreach to women who lack connectivity. This could be implemented through a peer group or other network intervention. In many developing countries, men are also more likely than women to own mobile phones. Therefore, it may be beneficial to involve male partners and community male leaders through mHealth initiatives to improve maternal care-seeking and knowledge. Finally, although women with limited mobile phone access were significantly less likely to recognize the importance of maternal health services and use them, knowledge and service utilization were relatively low regardless of mobile phone status. This underscores the importance of continued investment in multi-pronged strategies that leverage mobile and non-mobile maternal health service options. While mobile communication technologies have the potential to reduce maternal health disparities related to cost, distance, and inadequate health infrastructure, the technology alone is insufficient to address all of women’s health service needs. Given likewise the observed poorer health outcomes, women without mobile phones may be equally underserved
by conventional, non-mobile health service models. Therefore, increasing investments in integrated health delivery models that include facility- and community-based outreach with complementary, but not exclusive, mobile-based services may assist in minimizing health inequities brought on both by health systems and technological barriers in access to information and services.
4.1.
Limitations
The limitations of this study deserve mention. The study’s findings are based on a cross-sectional survey design. Therefore, the study was unable to investigate the causal relationship between mobile phone status and maternal careseeking and knowledge. Differences in outcomes may reflect the interventional impact of pre-existing mHealth initiatives, exposure to which was unmeasured in this analysis. However, given the social and geographic diversity of women included in the survey, it is unlikely that the findings reflect the impact of a single or set of mHealth interventions.
5.
Conclusions
As maternal mHealth strategies are increasingly utilized, preintervention assessments can provide important insights on the differences in maternal health needs by mobile phone status. Our findings suggest that in settings with unequal access to mobile phone technologies, mHealth interventions involving patient interaction are unlikely to reach women who have the poorest maternal knowledge and care-seeking as these women often lacked mobile connectivity. More efforts are needed to improve women’s access to mobile phones and minimize potential health inequities brought on by health systems and technological barriers in access to care.
Conflict of interests statement The authors declare that there are no conflicts of interest.
Authors’ contributions LJ conceived and designed study, developed the research instruments, analyzed the results, and wrote the first draft of the manuscript. AO oversaw data collection, supported the analysis, assisted in the interpretation of findings, and edited subsequent drafts of the manuscript. AA led all data management and cleaning, provided technical assistance in the statistical analyses, and made contributions to the manuscript. IY managed the evaluation and reviewed the manuscript’s findings and implications. EE supported the instrument development, field conduct, and manuscript writing.
Acknowledgements The authors wish to thank all the women and households who participated in the study, in addition to the data collectors, monitoring and evaluation advisors, and program staff of
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 4 ( 2 0 1 5 ) 341–348
Summary points What was already known on the topic: • Mobile communication technologies have the potential to reduce maternal health disparities related to cost, distance, and inadequate health infrastructure. • Despite expanding mobile phone coverage, significant disparities in access to mobile technology remain by gender, household, and geographic characteristics. • Implementing equitable maternal mHealth interventions requires information on women’s differential health needs by mobile phone strata. What this study added to our knowledge: • We assess whether women with limited mobile phone access have differential odds of maternal knowledge and health service utilization as compared to female mobile phone users currently eligible to participate in maternal mHealth programs. • Women with no mobile phone access were significantly less likely to utilize skilled obstetric services and recognize their benefits as compared to mobileconnected women. • In settings with unequal access to mobile phone technologies, mHealth interventions may not reach women who have the poorest maternal knowledge and care-seeking as these women often lacked access to mobile phones.
the PATHS2 project, the Nigeria Ministry of Health, the Nigeria National Bureau of Statistics, ITAD, Inc., and Abt Associates, Inc. who made this study possible. This work was supported by the (DFID) PATHS2 Project (Grant Number 200708080) at Abt Associates, Inc. and the Johns Hopkins Bloomberg School of Public Health. All conclusions are those of the authors and do not necessarily reflect the views of the funding or managing organizations.
references
[1] World Health Organization, Mhealth New horizons for health through mobile: based on findings of the second global survey on eHealth, Global Observ. eHealth Ser. 3 (2011) 1–122. [2] L. López, A.R. Green, A. Tan-McGrory, et al., Bridging the digital divide in health care: the role of health information technology in addressing racial and ethnic disparities, Jt. Comm. J. Qual. Patient Saf. 37 (10) (2011) 437–445. [3] World Bank, World Bank Information and Communications for Development 2012: Maximizing Mobile, World Bank, Washington, DC, 2012 (License: Creative Commons Attribution CC BY 3.0). [4] L. Newman, K. Biedrzycki, F. Baum, Digital technology use among disadvantaged Australians: implications for equitable consumer participation in digitally-mediated communication and information exchange with health services, Aust. Health Rev. 36 (2012) 125–129.
347
[5] N.M. Denizard-Thompson, K.B. Feiereisel, S.F. Stevens, et al., The digital divide at an urban community health center: implications for quality improvement and health care access, J. Community Health 36 (3) (2011) 456–460. [6] C. Ronquillo, L. Currie, The digital divide: trends in global mobile and broadband Internet access from 2000–2010, Nurs. Inf. (2012) 346. [7] Cherie Blair Foundation for Women, Women & Mobile: A Global Opportunity—A Study on the Mobile Phone Gender Gap in Low and Middle-income Countries, Cherie Blair Foundation for Women, 2010, Available at: http://www.gsma.com/mobilefordevelopment/ wpcontent/uploads/2013/01/GSMA Women and MobileA Global Opportunity.pdf. [8] J.E. Blumenstock, N. Eagle, Divided we call: disparities in access and use of mobile phones in Rwanda, Inf. Technol. Int. Dev. 8 (2) (2012) 1–16. [9] Z. Khadr, Monitoring socioeconomic inequity in maternal health indicators in Egypt: 1995–2005, Int. J. Equity Health 8 (38) (2008) 1–12. [10] P.K. Pathak, A. Singh, S.V. Subramanian, Economic inequalities in maternal health care: prenatal care and skilled birth attendance in India, 1992–2006, PLoS ONE 5 (10) (2010) e13593. [11] M. Wirth, E. Sacks, E. Delamonica, et al., Delivering on the MDGs?: equity and maternal health in Ghana, Ethiopia and Kenya, East Afr. J. Public Health 5 (3) (2008) 133–141. [12] J.T. Boerma, J. Bryce, Y. Kinfu, et al., Mind the gap: equity and trends in coverage of maternal, newborn, and child health services in 54 Countdown countries. Countdown 2008 Equity Analysis Group, Lancet 371 (2008) 1259–1267. [13] D. You, G. Jones, K. Hill, et al., Levels and trends in child mortality, 1990–2009, Lancet 376 (2010) 931–933. [14] M. Chopra, H. Campbell, I. Rudan, Understanding the determinants of the complex interplay between cost-effectiveness and equitable impact in maternal and child mortality reduction, J. Glob Health 2 (1) (2012) 10406. [15] Z. Quayyum, M.N.U. Khan, T. Quayyum, et al., Can community level interventions have an impact on equity and utilization of maternal health care–Evidence from rural Bangladesh, Int. J. Equity Health 12 (22) (2013) 1–13. [16] U. Sarkar, A.J. Karter, J.Y. Liu, et al., Social disparities in internet patient portal use in diabetes: evidence that the digital divide extends beyond access, J. Am. Med. Inform. Assoc. 18 (3) (2011) 318–321. [17] P. Braveman, S. Gruskin, Defining equity in health, J. Epidemiol. Comm. Health 57 (4) (2003) 254–258. [18] S.I. Wang, C.L. Yaung, Vertical equity of healthcare in Taiwan: health services were distributed according to need, Int. J. Equity Health 31;12 (2013) 12. [19] World Health Organization, Trends in Maternal Mortality: 1990 to 2010, in: WHO, UNICEF, UNFPA and The World Bank Estimates, World Health Organization, 2012, Available at URL: https://www.unfpa.org/webdav/site/global/shared/ documents/publications/2012/Trends in maternal mortality A4-1.pdf. [20] World Health Organization, Maternal Mortality in 1990–2013, in: WHO, UNICEF, UNFPA, The World Bank, and United Nations Population Division Maternal Mortality Estimation Inter-Agency Group, Global Health Observatory—Maternal, Neonatal and Child Health, Nigeria, 2013, Available at URL: http://www.who.int/gho/maternal health/countries/nga. pdf?ua=1. [21] World Health Organization, Standards of maternal and neonatal care, in: Integrated Management of Pregnancy and Childbirth, Department of Making Pregnancy Safer, 2007, Available at URL: http://whqlibdoc.who.int/hq/2007/ a91272.pdf.
348
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 4 ( 2 0 1 5 ) 341–348
[22] Nigeria Demographic Health Survey, Preliminary Report. National Population Commission, MEASURE DHS/ICF International, Abuja, Nigeria/Calverton, MD, 2013, Available at URL: http://dhsprogram.com/pubs/pdf/PR41/PR41.pdf. [23] O. Lincetto, S. Mothebesoane-Anoh, P. Gomez, et al., Antenatal care: Chapter 2, in: Opportunities for Africa’s Newborns, 2006, http://www.who.int/pmnch/media/publications/ aonsectionIII 2.pdf. [24] Nigeria Communications Commission, Industry Information: Subscriber Statistics—Monthly Subscriber Data, Nigeria Communications Commission, April 2014, URL: http://www.ncc.gov.ng. [25] W.M. Olatokun, Availability, accessibility, and use of ICTs by Nigerian women academics. Africa Regional Centre for Information Science (ARCIS), Malays. J. Lib. Info Sci. 12 (2) (2007) 13–33. [26] Mobile broadband for working women: Nigeria. 2013. Transforming women’s livelihoods through mobile
[27]
[28]
[29]
[30]
broadband. QualComm. mWomen. Available at URL: https://www.qualcomm.com/documents/transformingwomens-livelihoods-through-mobile-broadband. H.V. Doctor, A. Olatunji, A. Jumare, Bridging the communication gap: successes and challenges of mobile phone technology in a health and demographic surveillance system in northern Nigeria, Online J. Public Health Inf. 4 (3) (2012) e14. C.B. Aranda-Jan, N. Mohutsiwa-Dibe, S. Loukanova, Systematic review of what works, what does not work and why of implementation of mobile health (mHealth) projects in Africa, BMC Public Health 14 (2014) 188. T.J. Betjeman, S.E. Soghoian, M.P. Foran, Review article: mHealth in Sub-Saharan Africa, Intl. J. Telemed Appl. (2013) e482324. A. Chib, H. Wilkin, B. Hoefman, Vulnerabilities in mHealth implementation: an Uganda HIV/AIDS SMS campaign, Global Health Prom. 20 (1 Suppl) (2013) 26–32.
journal homepage: www.ijmijournal.com
Disparities in mobile phone access and maternal health service utilization in Nigeria: A population-based survey Larissa Jennings a,b,∗ , Adetayo Omoni b,c , Akunle Akerele c , Yisa Ibrahim c , Ekpenyong Ekanem b,c a
Johns Hopkins Bloomberg School of Public Health, Department of International Health, 615N. Wolfe Street, Room E5038, Baltimore, MD 21205, USA b Abt Associates, Inc., International Health Division, Monitoring and Evaluation, 4550 Montgomery Ave, Suite 800N, Bethesda, MD 20814, USA c Partnership for Transforming Health Systems II (PATH2), Monitoring and Evaluation, 37 Panama Street, IBB Way, Maitama, Abuja, Nigeria
a r t i c l e
i n f o
a b s t r a c t
Article history:
Background: Mobile communication technologies may reduce maternal health disparities
Received 11 July 2014
related to cost, distance, and infrastructure. However, the ability of mHealth initiatives to
Received in revised form
accelerate maternal health goals requires in part that women with the greatest health needs
14 January 2015
have access to mobile phones.
Accepted 22 January 2015
Objective: This study examined if women with limited mobile phone access have differential odds of maternal knowledge and health service utilization as compared to female mobile
Keywords:
phone users who are currently eligible to participate in maternal mHealth programs.
Mobile phones
Methods: Using household survey data from Nigeria, multivariable logistic regressions were
Maternal health
used to examine the odds of maternal knowledge and service utilization by mobile phone
Service utilization
strata.
Access
Results: Findings showed that in settings with unequal access to mobile phones, mHealth
Equity
interventions may not reach women who have the poorest maternal knowledge and care-
Disparities
seeking as these women often lacked mobile connectivity. As compared to mobile users, women without mobile phone access had significantly lower odds of antenatal care utilization (OR = 0.48, 95%CI: 0.36–0.64), skilled delivery (OR = 0.56, 95%CI: 0.45–0.70), and modern contraceptive use (OR = 0.50, 95%CI: 0.33–0.76) after adjusting for demographic characteristics. They also had significantly lower knowledge of maternal danger signs (OR = 0.69, 95%CI: 0.53–0.90) and knowledge of antenatal (OR = 0.46, 95%CI: 0.36–0.59) and skilled delivery care benefits (OR = 0.62, 95%CI: 0.47–0.82). No differences were observed by mobile phone strata in uptake of emergency obstetric care, postnatal services, or breastfeeding. Conclusions: As maternal mHealth strategies are increasingly utilized, more efforts are needed to improve women’s access to mobile phones and minimize potential health inequities brought on by health systems and technological barriers in access to care. © 2015 Elsevier Ireland Ltd. All rights reserved.
∗
Corresponding author at: Johns Hopkins Bloomberg School of Public Health, Department of International Health, 615 N. Wolfe Street, E5038, Baltimore, MD 21205, USA. Tel.: +1 410 955 3537. E-mail addresses: [email protected], [email protected] (L. Jennings), adetayo [email protected] (A. Omoni), [email protected] (A. Akerele), [email protected] (Y. Ibrahim), ekpenyong [email protected] (E. Ekanem). http://dx.doi.org/10.1016/j.ijmedinf.2015.01.016 1386-5056/© 2015 Elsevier Ireland Ltd. All rights reserved.
342
1.
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 4 ( 2 0 1 5 ) 341–348
Introduction
Mobile communication technologies have the potential to reduce maternal health disparities related to cost, distance, and inadequate health infrastructure, even in technologypoor settings [1–3]. Yet, the ability of mobile health (mHealth) initiatives to accelerate maternal development goals requires in part that women with the greatest health needs have access to mobile communication devices [4,5]. While mobile phone coverage in developing countries has rapidly increased in recent years [3,6], research has shown that significant disparities in access remain by gender, household, and geographic characteristics [4,7,8]. Women account for almost two-thirds of the global mobile phone market that has not yet been reached [7], and those who are the poorest and least educated are most often lacking connectivity [3,4,7]. Poor and uneducated women also have higher rates of prenatal, intranatal, and postnatal morbidity and mortality [9–11]. While substantial gains have been made in addressing maternal morbidity and mortality, some achievements have resulted historically in greater health inequity at the population level with significantly lower uptake among the most vulnerable and marginalized women [12–15]. This has raised concerns whether the expansion of maternal mHealth applications that directly target patients-such as public health communications, personalized behavior change messages, or self-care information is likely to exacerbate health disparities among women most at-risk for adverse health outcomes [2,4,16]. Implementing equitable maternal mHealth interventions requires information on women’s differential health needs by mobile phone strata [3–5,13,17,18]. However, few studies have examined need-related inequalities in maternal service utilization as compared to mobile phone access. This has limited the ability of mobile-based maternal health programs to assess beforehand the potential of serving women and families most at-risk. Given that mHealth applications are increasingly being utilized to improve the delivery of maternal health services, this analysis aimed to preliminarily inform the development of a community-based mHealth initiative targeting women with low maternal knowledge and care-seeking behaviors. We sought to determine whether a mHealth intervention aiming to increase uptake of maternal services through participant interaction was likely to reach women with the highest unmet maternal need (i.e., low antenatal use, unskilled delivery attendance, poor maternal knowledge, etc.), as indicated by access to a mobile phone. Specifically, we examined if women with limited mobile phone access had differential odds of maternal health service utilization or knowledge of pregnancy, birth, and postpartum health as compared to female mobile phone users who are currently eligible to participate in maternal mHealth programs. Implications on equitable implementation of mobile technology to reach the world’s most vulnerable women are discussed.
2.
Methods
2.1.
Sampling
Data were used from the Partnership for Transforming Health Systems, Phase Two (PATHS2) project’s baseline household
survey in Nigeria. Using a population-based, two-stage cluster design, the PATHS2 household survey was conducted in five project states: Enugu, Jigawa, Kaduna, Kano, and Lagos to measure pre-intervention prevalence of a range of maternal, reproductive, and child health outcomes. In the first stage, 625 enumeration areas were selected from the local government areas within the five states. In the second stage, a target sample of 10,000 households was selected from the enumeration areas. Local and trained interviewers then surveyed the head of household and, if applicable, one woman aged 15–49 with a child 23 months or less was randomly selected from the household.
2.2.
Maternal health setting
Despite declines in recent years, Nigeria accounts for an estimated 14% of maternal deaths worldwide, the second largest contributor to maternal mortality in the world [19]. In 2013, an estimated 40,000 Nigerian women died in pregnancy and birth-related complications representing approximately 560 maternal deaths for every 100,000 live births [20]. Care from a skilled provider during pregnancy and childbirth can reduce the risk of complications and infections that lead to illness or death of mothers and newborns [21–23]. Yet, only 61% of women in Nigeria reported receiving antenatal care from a trained provider during their most recent pregnancy, and only 38% of Nigerian births were delivered by a skilled health provider [22]. As of 2014, there were an estimated 129 million mobile phones in use in Nigeria, representing about 76% of the total population [24]. However, mobile phone coverage varies substantially across the country, and women are significantly less likely than men to own a mobile phone [25–27].
2.3.
Data collection
Data were collected in July and August 2012. A structured questionnaire was used to elicit information on household demographics and women’s maternal and reproductive histories. The questionnaire was developed in English and blind, back-translated into Yoruba, Igbo, and Hausa, the languages spoken predominantly in the study areas. All survey questions were administered orally using personal digital assistants in a language chosen by the participant. Interviewers were trained on the survey’s objective and methodology using didactic methods as well as mock interviews and pilot tests.
2.4.
Outcome measures
The outcome measures for the study comprised maternal knowledge and health service utilization during three periods: pregnancy, labor and delivery, and the early postnatal period. Two care-seeking behaviors were assessed during pregnancy. Women were asked if they received antenatal care during their last pregnancy: yes (code = 1) or no (code = 0). For women who received antenatal care, we also dichotomized the sample among those who received the minimum recommended four antenatal care visits (code = 1) versus those who received three or fewer visits (code = 0). To assess maternal knowledge regarding care during pregnancy, women were asked to describe the advantages of attending antenatal care.
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 4 ( 2 0 1 5 ) 341–348
Responses were matched against five pre-coded antenatal care benefits: receipt of health education, examination by a skilled provider, early detection and management of complications, receipt of drugs/medicines, malaria prevention, and anemia prevention. An antenatal benefit corresponding to one of the five responses generated a score of 1, where a knowledge score ≥ 3 was coded as 1 compared to a knowledge score < 3 (code = 0). Women were also asked what danger signs would indicate that a woman who is pregnant, in labor, or has recently delivered needs to be rushed to a health facility for emergency care. Women who knew of at least three maternal danger signs (severe headache, swelling, dizziness, fever/chills, convulsions, hemorrhage, prolonged labor, malpresentation, or retained placenta) were coded as 1 compared to women who were less knowledgeable, coded as 0. Four health care-seeking scenarios were examined pertaining to labor and delivery. Women were asked who attended the delivery of their last child: a doctor, nurse/midwife, or auxiliary midwife (code = 1) versus those with unskilled assistance (code = 0). Women reported likewise if they experienced any complication during their last child’s birth (severe headache, convulsions, hemorrhage, prolonged labor, mal-presentation, or retained placenta): yes (code = 1) or no (code = 0). If during labor and delivery of their last child an ambulatory transportation service was contacted for emergency travel to a health center, women were coded as yes (code = 1) or no (code = 0). To examine maternal knowledge regarding delivery care, women were asked to give two reasons why delivering in a health facility was beneficial: ≥ 2 responses out of the nine precoded responses (receipt of health education, management of obstetric complications, prevention of infection, prevention of vertical HIV transmission, immediate newborn care, provision of other necessary treatment/drugs, avoidance of emergency care delays, health provider assistance, and psychosocial support) were coded as 1, compared to < 2 responses, coded as 0. Two questions on postnatal care utilization examined whether a doctor, nurse/midwife, or health center staff visited the woman at home in the first week after her most recent birth: yes (code = 1) or no (code = 0). If yes, the woman was also asked how many days after birth did a skilled provider first visit her home: within a day (code = 1) or within 2 to 7 days (code = 0). Women were also asked if they ever breastfed their youngest child, and if yes, how long after birth did they initiate breastfeeding: within the first hour of birth (code = 1) versus > 1 hour (code = 0). Lastly, all women indicated whether they were currently using any method to delay or avoid pregnancy: modern contraceptive method (code = 1) versus those who were not (code = 0). Information was also obtained on each woman’s age, highest level of education attained, location (rural/urban), state, and household wealth index.
2.5.
Exposure measure
The exposure variable of interest was mobile phone status. Women were asked if they currently had access to a mobile phone, defined as having the option to use in a singular or shared fashion a functional cellular phone. Women who responded “yes” were categorized as the reference group
343
(code = 0) compared to women who did not have mobile phone access (code = 1).
2.6.
Data analysis
Data were analyzed using STATA, Version 12 (Stata Corporation, College Station, TX). Analyses were done with sampling weights using the svy command in STATA to account for the survey cluster design. We first examined the distribution of sample demographic characteristics and outcomes by mobile phone status using two-sample tests of proportions and chisquare (2 ) statistics. Next, bivariate and multivariable logistic regression models were used to examine differences in the odds of maternal knowledge and service utilization by mobile phone status. Odds ratios were considered statistically significant at p < .05, when the null value of 1.0 was not included in the corresponding 95% confidence interval.
3.
Results
3.1.
Sample characteristics
A total of 3,390 women aged 15–49 with a child 23 months or less were identified by the survey (Table 1) out of a final sample of 10,107 households. The mean age of women was 28.3 years, and 51.3% had received primary education or higher. Nearly two-thirds (61.3%) of women lived in a rural area, and those from Kano (33.8%) and Lagos (24.1%) states made up the largest proportions of the total sample. Mobile phone access varied by demographic characteristics. Approximately half (51.4%) of women did not have access to a mobile phone. As compared to those who reported having mobile phone access, women who did not were significantly younger (27.2 years vs. 29.4 years, p < 0.001), had lower rates of primary educational attainment (27.2% vs. 66.1%, p < 0.001), and were more likely to live in a rural setting (82.5% vs. 37.0%, p < 0.001). Women without access to a mobile phone also made up greater proportions of the poorer and poorest household wealth quintiles (72.0% vs. 16.5%, p < 0.001).
3.2.
Antenatal care
Seventy-five percent (75.0%) of women received antenatal care by a skilled provider at least once during their most recent pregnancy, and 69.1% received all four recommended antenatal visits (Table 2). Antenatal care utilization varied considerably by mobile phone status. In bivariate models, women without access to a mobile phone had 83% significantly lower odds (OR = 0.17, 95%CI: 0.14–0.22) of receiving any skilled antenatal care, and 53% significantly lower odds (OR = 0.47, 95%CI: 0.38–0.58) of having at least four skilled antenatal care visits (Table 3). These differences remained after controlling for demographic characteristics. Compared to mobile phone users, the adjusted odds of attending one or all of the recommended four antenatal care visits was 52% lower (OR = 0.48, 95%CI: 0.36–0.64) and 29% lower (OR = 0.71, 95%CI: 0.55–0.92), respectively, among women without access to a mobile phone.
344
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 4 ( 2 0 1 5 ) 341–348
Table 1 – Demographic characteristics for total sample and by mobile phone access. By Mobile Phone Access
Total
Frequency (n = ) Proportion of sampled womena (%) Maternal age in years (mean ± SE) Education Primary or more Less than primary Location Rural Urban State Enugub Jigawa Kaduna Kano Lagos Household wealth quintile Poorest Poorer Middle Richer Richest a b ∗
3930 100.0 28.3 (±0.1)
With
Without
p-Value
1811 46.8 29.4 (±0.2)
1998 51.4 27.2 (±0.2)
– – 0.000*
51.3 43.5
76.9 19.7
27.2 66.1
0.000*
61.3 38.7
37.0 63.0
82.5 17.5
0.000*
5.9 21.9 14.4 33.8 24.1
8.2 13.6 11.6 20.3 46.4
0.9 30.2 17.3 47.1 4.6
0.000*
23.8 22.5 17.6 18.4 17.6
6.4 11.1 17.5 29.4 35.6
39.9 32.9 17.1 8.6 1.6
0.000*
Population estimates are weighted to adjust for cluster survey design. Phone access data missing for 26.8% of sample. Significant at p < 0.05.
Twenty-seven percent (27.2%) of women knew three or more benefits to antenatal care (Table 2). Even fewer (18.9%) were knowledgeable of at least three maternal danger signs requiring urgent care. After controlling for demographic characteristics, women without mobile phone access had 54% significantly lower odds (OR = 0.46, 95%CI: 0.36–0.59) than female mobile phone users of knowing the benefits of antenatal care (Table 3). In multivariable analyses, women lacking mobile connectivity also had 31% significantly decreased odds
(OR = 0.69, 95%CI: 0.53–0.90) of correctly identifying pregnancy danger signs compared to mobile-connected women.
3.3.
Labor and delivery
Similar patterns were observed in labor and delivery care utilization by mobile phone status. While nearly half (48.3%) of women delivered with a skilled health provider (Table 2), women who lacked access to a mobile device had an 82%
Table 2 – Maternal service utilization and knowledge for total sample and by mobile phone access. Proportion of sampled women (%)
Frequency (N=) Antenatal care Received at least 1 antenatal care visit Received recommended 4 antenatal care visits Maternal knowledge of benefits of antenatal care Maternal knowledge of obstetric danger signs Labor and delivery Delivered with skilled birth attendant Had obstetric complication Used ambulatory transportation for obstetric emergency Maternal knowledge of benefits of skilled birth attendance Postnatal care Received at least one postnatal visit within one week Received at least one postnatal visit within day of birth Initiation of breastfeeding within one hour of birth Uptake of modern contraception a ∗
Population estimates are weighted to adjust for cluster survey design. Significant at p < .05
By Mobile Phone Access
Total With
Without
1811
1998
75.0 69.1 27.2 18.9
90.1 76.2 38.7 22.8
61.2 60.0 16.7 15.3
0.000* 0.000* 0.000* 0.000*
48.3 10.3 3.7 74.9
69.2 11.3 5.1 86.1
28.7 9.5 2.3 64.8
0.000* 0.141 0.000* 0.000*
10.1 32.7 40.8 7.5
12.3 31.9 37.9 12.4
8.1 33.9 43.6 3.1
0.002* 0.759 0.012a,* 0.000*
3930
p-Value –
345
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 4 ( 2 0 1 5 ) 341–348
Table 3 – Odds ratios and 95% confidence intervals of maternal service utilization for women without mobile phone access compared to reference group. Bivariate modela,c
Total: N = 3930 OR Antenatal care Received at least 1 antenatal care visit Received recommended 4 antenatal care visits Maternal knowledge of benefits of antenatal care Maternal knowledge of obstetric danger signs Labor and delivery Delivered with skilled birth attendant Had obstetric complication Used ambulatory transportation for obstetric emergency Maternal knowledge of benefits of skilled birth attendance Postnatal care Received at least 1 PNC visit within one week Received at least 1 PNC visit within day of birth Initiation of breastfeeding with one hour of birth Uptake of modern contraception a b
c ∗
95% CI
Multivariable modelb,c OR
95% CI
0.17* 0.47* 0.32* 0.61*
0.14–0.22 0.38–0.58 0.26–0.39 0.49–0.75
0.48* 0.71* 0.46* 0.69*
0.36–0.64 0.55–0.92 0.36–0.59 0.53–0.90
0.18* 0.82 0.44* 0.30*
0.15–0.22 0.64–1.07 0.29–0.68 0.24–0.37
0.56* 0.80 0.82 0.62*
0.45–0.70 0.58–1.11 0.44–1.53 0.47–0.82
0.63* 1.09 1.27* 0.23*
0.48–0.84 0.62–1.92 1.06–1.53 0.16–0.32
1.21 0.72 1.02 0.50*
0.83–1.76 0.35–1.48 0.81–1.27 0.33–0.76
Bivariate model adjusted for cluster survey design only. Multivariable model adjusted for cluster survey design and control variables [maternal age, education, location (rural/urban), state, and household wealth] with mean variance inflation factors ranging: 1.27 to 1.85. Index group = women without mobile phone access (code = 1); Reference group = women with mobile phone access (code = 0). Significant at p < .05.
significant decrease (OR = 0.18, 95%CI: 0.15–0.22) in the odds of delivering with a skilled attendant as compared to those with mobile phone access (Table 3). This relationship was maintained at 44% significantly lower odds (OR = 0.56, 95%CI: 0.45–0.70) of delivering with a skilled provider after controlling for demographic factors. Most surveyed women (74.9%) were knowledgeable of at least two benefits to having skilled delivery care (Table 2), but women without mobile phone access were 38% significantly less likely (OR = 0.62, 95%CI: 0.47–0.82) to be aware as compared to their mobile counterparts in multivariable models (Table 3). In contrast, no significant differences were observed in multivariable models in the proportion of women who had a complication during childbirth by mobile phone status (OR = 0.80, 95%CI: 0.58–1.11), nor in ambulatory transportation use for obstetric emergencies (OR = 0.82, 95%CI: 0.44–1.53).
3.4.
Postnatal care
Postnatal care utilization was lower among surveyed women than antenatal care utilization. A third (32.7%) of women received at least one postnatal visit at home within one day of birth, and 10.1% within the first week after delivery (Table 2). No significant differences were observed in multivariable models in the odds of postnatal care utilization by phone status within a day of birth (OR = 0.72, 95%CI: 0.35–1.48) or within one week following birth (OR = 1.21, 95%CI: 0.83–1.76) (Table 3). Among postnatal care practices, 40.8% of women reported initiating breastfeeding within the first hour of birth, and women lacking mobile connectivity had a 1.27 times significantly greater odds of immediate breastfeeding (OR = 1.27 95%CI: 1.06–1.53). However, this association was not statistically significant in multivariable analyses (OR = 1.02, 95%CI: 0.81–1.27). Uptake of modern contraceptive methods was low (7.5%) among all women, but women who did not
have access to a mobile phone had a significant 50% decrease (OR = 0.50, 95%CI: 0.33–0.76) in the odds of using any modern contraceptive method, after controlling for demographic characteristics.
4.
Discussion
The ability of mHealth initiatives to accelerate maternal development goals requires in part that women with the greatest health care needs have access to mobile devices. Yet, few studies have examined maternal health priorities by mobile phone status. This study found that after controlling for demographic factors, women with no mobile phone access were significantly less likely to utilize skilled antenatal and delivery care services as compared to mobile-connected women who are currently eligible to participate in patient-based maternal mHealth interventions. Women without access to mobile phones were also significantly less knowledgeable of the benefits of skilled maternal care and dangers signs requiring urgent medical attention. Such findings suggest that in settings with unequal access to mobile phone technologies, mHealth programs that aim to increase uptake of maternal services through direct support to patients may be less effective in reaching women with the poorest maternal knowledge and care-seeking as these women often lacked access to mobile phones. Other studies have also highlighted this weakness [28]. A recent review of mHealth interventions in sub-Saharan Africa found that mHealth initiatives aiming to disseminate health education had been less successful due to limited accessibility among the most vulnerable groups [29]. In Uganda, evaluators of an HIV/AIDS text-messaging campaign indicated also that the intervention failed to provide information to the most vulnerable women, those with the lowest income and who did not own mobile phones [30].
346
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 4 ( 2 0 1 5 ) 341–348
Given the growing need to ensure that the benefits of maternal mHealth strategies are equitably distributed, this paper offers several considerations for enhancing future practice and design. One important strategy for mHealth implementers who are targeting the most vulnerable women is to assess during the development phase the potential distributional impact of mobile-based strategies on maternal outcomes. Careful planning is needed in determining whether the intended mobile technology is accessible to women with the greatest maternal health needs. While the process of measuring equitable impact is complex [14], the results from this study suggest that a simple, pre-intervention assessment can provide important insights during programmatic design on differences in maternal health needs by mobile phone status. Such information may be used by implementers to tailor maternal mHealth programs with technologicallyappropriate components for women with varying levels of mobile connectivity. A second implication is to improve women’s access to mobile phones, since women who lacked mobile access in this study represented the population most likely to benefit from improved maternal health interventions. Research has shown that expanding access to mobile communication devices among women in resource-poor settings can yield several social, economic, and health benefits [7]. This may reflect empowerment differences in women who were able to access and utilize new technologies. On the other hand, empowering women by improving access to mobile technologies may strengthen their capacity to utilize maternal health-related services, including skilled antenatal and delivery care, and increase their participation in social and economic activities that positively impact reproductive decision-making. Results from this study suggest that the ubiquity of mobile phones is not universal. Greater numbers of women will likely need access to mobile phones in order to improve the capacity of mHealth strategies to minimize rather than exacerbate maternal health disparities in resource-poor settings. In contrast, given that roughly half of the surveyed women had access to mobile phones, this study also highlights the possibility of engaging current female mobile phone users as transfer agents of mobile-based outreach to women who lack connectivity. This could be implemented through a peer group or other network intervention. In many developing countries, men are also more likely than women to own mobile phones. Therefore, it may be beneficial to involve male partners and community male leaders through mHealth initiatives to improve maternal care-seeking and knowledge. Finally, although women with limited mobile phone access were significantly less likely to recognize the importance of maternal health services and use them, knowledge and service utilization were relatively low regardless of mobile phone status. This underscores the importance of continued investment in multi-pronged strategies that leverage mobile and non-mobile maternal health service options. While mobile communication technologies have the potential to reduce maternal health disparities related to cost, distance, and inadequate health infrastructure, the technology alone is insufficient to address all of women’s health service needs. Given likewise the observed poorer health outcomes, women without mobile phones may be equally underserved
by conventional, non-mobile health service models. Therefore, increasing investments in integrated health delivery models that include facility- and community-based outreach with complementary, but not exclusive, mobile-based services may assist in minimizing health inequities brought on both by health systems and technological barriers in access to information and services.
4.1.
Limitations
The limitations of this study deserve mention. The study’s findings are based on a cross-sectional survey design. Therefore, the study was unable to investigate the causal relationship between mobile phone status and maternal careseeking and knowledge. Differences in outcomes may reflect the interventional impact of pre-existing mHealth initiatives, exposure to which was unmeasured in this analysis. However, given the social and geographic diversity of women included in the survey, it is unlikely that the findings reflect the impact of a single or set of mHealth interventions.
5.
Conclusions
As maternal mHealth strategies are increasingly utilized, preintervention assessments can provide important insights on the differences in maternal health needs by mobile phone status. Our findings suggest that in settings with unequal access to mobile phone technologies, mHealth interventions involving patient interaction are unlikely to reach women who have the poorest maternal knowledge and care-seeking as these women often lacked mobile connectivity. More efforts are needed to improve women’s access to mobile phones and minimize potential health inequities brought on by health systems and technological barriers in access to care.
Conflict of interests statement The authors declare that there are no conflicts of interest.
Authors’ contributions LJ conceived and designed study, developed the research instruments, analyzed the results, and wrote the first draft of the manuscript. AO oversaw data collection, supported the analysis, assisted in the interpretation of findings, and edited subsequent drafts of the manuscript. AA led all data management and cleaning, provided technical assistance in the statistical analyses, and made contributions to the manuscript. IY managed the evaluation and reviewed the manuscript’s findings and implications. EE supported the instrument development, field conduct, and manuscript writing.
Acknowledgements The authors wish to thank all the women and households who participated in the study, in addition to the data collectors, monitoring and evaluation advisors, and program staff of
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 4 ( 2 0 1 5 ) 341–348
Summary points What was already known on the topic: • Mobile communication technologies have the potential to reduce maternal health disparities related to cost, distance, and inadequate health infrastructure. • Despite expanding mobile phone coverage, significant disparities in access to mobile technology remain by gender, household, and geographic characteristics. • Implementing equitable maternal mHealth interventions requires information on women’s differential health needs by mobile phone strata. What this study added to our knowledge: • We assess whether women with limited mobile phone access have differential odds of maternal knowledge and health service utilization as compared to female mobile phone users currently eligible to participate in maternal mHealth programs. • Women with no mobile phone access were significantly less likely to utilize skilled obstetric services and recognize their benefits as compared to mobileconnected women. • In settings with unequal access to mobile phone technologies, mHealth interventions may not reach women who have the poorest maternal knowledge and care-seeking as these women often lacked access to mobile phones.
the PATHS2 project, the Nigeria Ministry of Health, the Nigeria National Bureau of Statistics, ITAD, Inc., and Abt Associates, Inc. who made this study possible. This work was supported by the (DFID) PATHS2 Project (Grant Number 200708080) at Abt Associates, Inc. and the Johns Hopkins Bloomberg School of Public Health. All conclusions are those of the authors and do not necessarily reflect the views of the funding or managing organizations.
references
[1] World Health Organization, Mhealth New horizons for health through mobile: based on findings of the second global survey on eHealth, Global Observ. eHealth Ser. 3 (2011) 1–122. [2] L. López, A.R. Green, A. Tan-McGrory, et al., Bridging the digital divide in health care: the role of health information technology in addressing racial and ethnic disparities, Jt. Comm. J. Qual. Patient Saf. 37 (10) (2011) 437–445. [3] World Bank, World Bank Information and Communications for Development 2012: Maximizing Mobile, World Bank, Washington, DC, 2012 (License: Creative Commons Attribution CC BY 3.0). [4] L. Newman, K. Biedrzycki, F. Baum, Digital technology use among disadvantaged Australians: implications for equitable consumer participation in digitally-mediated communication and information exchange with health services, Aust. Health Rev. 36 (2012) 125–129.
347
[5] N.M. Denizard-Thompson, K.B. Feiereisel, S.F. Stevens, et al., The digital divide at an urban community health center: implications for quality improvement and health care access, J. Community Health 36 (3) (2011) 456–460. [6] C. Ronquillo, L. Currie, The digital divide: trends in global mobile and broadband Internet access from 2000–2010, Nurs. Inf. (2012) 346. [7] Cherie Blair Foundation for Women, Women & Mobile: A Global Opportunity—A Study on the Mobile Phone Gender Gap in Low and Middle-income Countries, Cherie Blair Foundation for Women, 2010, Available at: http://www.gsma.com/mobilefordevelopment/ wpcontent/uploads/2013/01/GSMA Women and MobileA Global Opportunity.pdf. [8] J.E. Blumenstock, N. Eagle, Divided we call: disparities in access and use of mobile phones in Rwanda, Inf. Technol. Int. Dev. 8 (2) (2012) 1–16. [9] Z. Khadr, Monitoring socioeconomic inequity in maternal health indicators in Egypt: 1995–2005, Int. J. Equity Health 8 (38) (2008) 1–12. [10] P.K. Pathak, A. Singh, S.V. Subramanian, Economic inequalities in maternal health care: prenatal care and skilled birth attendance in India, 1992–2006, PLoS ONE 5 (10) (2010) e13593. [11] M. Wirth, E. Sacks, E. Delamonica, et al., Delivering on the MDGs?: equity and maternal health in Ghana, Ethiopia and Kenya, East Afr. J. Public Health 5 (3) (2008) 133–141. [12] J.T. Boerma, J. Bryce, Y. Kinfu, et al., Mind the gap: equity and trends in coverage of maternal, newborn, and child health services in 54 Countdown countries. Countdown 2008 Equity Analysis Group, Lancet 371 (2008) 1259–1267. [13] D. You, G. Jones, K. Hill, et al., Levels and trends in child mortality, 1990–2009, Lancet 376 (2010) 931–933. [14] M. Chopra, H. Campbell, I. Rudan, Understanding the determinants of the complex interplay between cost-effectiveness and equitable impact in maternal and child mortality reduction, J. Glob Health 2 (1) (2012) 10406. [15] Z. Quayyum, M.N.U. Khan, T. Quayyum, et al., Can community level interventions have an impact on equity and utilization of maternal health care–Evidence from rural Bangladesh, Int. J. Equity Health 12 (22) (2013) 1–13. [16] U. Sarkar, A.J. Karter, J.Y. Liu, et al., Social disparities in internet patient portal use in diabetes: evidence that the digital divide extends beyond access, J. Am. Med. Inform. Assoc. 18 (3) (2011) 318–321. [17] P. Braveman, S. Gruskin, Defining equity in health, J. Epidemiol. Comm. Health 57 (4) (2003) 254–258. [18] S.I. Wang, C.L. Yaung, Vertical equity of healthcare in Taiwan: health services were distributed according to need, Int. J. Equity Health 31;12 (2013) 12. [19] World Health Organization, Trends in Maternal Mortality: 1990 to 2010, in: WHO, UNICEF, UNFPA and The World Bank Estimates, World Health Organization, 2012, Available at URL: https://www.unfpa.org/webdav/site/global/shared/ documents/publications/2012/Trends in maternal mortality A4-1.pdf. [20] World Health Organization, Maternal Mortality in 1990–2013, in: WHO, UNICEF, UNFPA, The World Bank, and United Nations Population Division Maternal Mortality Estimation Inter-Agency Group, Global Health Observatory—Maternal, Neonatal and Child Health, Nigeria, 2013, Available at URL: http://www.who.int/gho/maternal health/countries/nga. pdf?ua=1. [21] World Health Organization, Standards of maternal and neonatal care, in: Integrated Management of Pregnancy and Childbirth, Department of Making Pregnancy Safer, 2007, Available at URL: http://whqlibdoc.who.int/hq/2007/ a91272.pdf.
348
i n t e r n a t i o n a l j o u r n a l o f m e d i c a l i n f o r m a t i c s 8 4 ( 2 0 1 5 ) 341–348
[22] Nigeria Demographic Health Survey, Preliminary Report. National Population Commission, MEASURE DHS/ICF International, Abuja, Nigeria/Calverton, MD, 2013, Available at URL: http://dhsprogram.com/pubs/pdf/PR41/PR41.pdf. [23] O. Lincetto, S. Mothebesoane-Anoh, P. Gomez, et al., Antenatal care: Chapter 2, in: Opportunities for Africa’s Newborns, 2006, http://www.who.int/pmnch/media/publications/ aonsectionIII 2.pdf. [24] Nigeria Communications Commission, Industry Information: Subscriber Statistics—Monthly Subscriber Data, Nigeria Communications Commission, April 2014, URL: http://www.ncc.gov.ng. [25] W.M. Olatokun, Availability, accessibility, and use of ICTs by Nigerian women academics. Africa Regional Centre for Information Science (ARCIS), Malays. J. Lib. Info Sci. 12 (2) (2007) 13–33. [26] Mobile broadband for working women: Nigeria. 2013. Transforming women’s livelihoods through mobile
[27]
[28]
[29]
[30]
broadband. QualComm. mWomen. Available at URL: https://www.qualcomm.com/documents/transformingwomens-livelihoods-through-mobile-broadband. H.V. Doctor, A. Olatunji, A. Jumare, Bridging the communication gap: successes and challenges of mobile phone technology in a health and demographic surveillance system in northern Nigeria, Online J. Public Health Inf. 4 (3) (2012) e14. C.B. Aranda-Jan, N. Mohutsiwa-Dibe, S. Loukanova, Systematic review of what works, what does not work and why of implementation of mobile health (mHealth) projects in Africa, BMC Public Health 14 (2014) 188. T.J. Betjeman, S.E. Soghoian, M.P. Foran, Review article: mHealth in Sub-Saharan Africa, Intl. J. Telemed Appl. (2013) e482324. A. Chib, H. Wilkin, B. Hoefman, Vulnerabilities in mHealth implementation: an Uganda HIV/AIDS SMS campaign, Global Health Prom. 20 (1 Suppl) (2013) 26–32.
