Mobile health technology to improve maternal health awareness in tribal populations: mobile for mothers

Abstract

Mobile health (mHealth) applications have the potential to improve health awareness. This study reports a quasi-controlled intervention to augment maternal health awareness among tribal pregnant mothers through the mHealth application. Households from 2 independent villages with similar socio-demographics in tribal regions of India were selected as intervention (Village A) and control group (Village B). The control group received government mandated programs through traditional means (orally), whereas the intervention group received the same education through mHealth utilization. Postintervention, awareness about tetanus injections and consumption of iron tablets was significantly (P < .001) improved in the intervention group by 55% and 58%, respectively. Awareness about hygiene significantly (P < .001) increased by 57.1%. In addition, mothers in the intervention group who recognized vaginal bleeding, severe abdominal pain, severe blurring of vision, or convulsions as danger signs during pregnancy significantly (P < .001) increased by 18.30%, 23.2%, 20.0%, and 4.90%, respectively. Our study indicates that despite the low literacy of users, mHealth intervention can improve maternal health awareness among tribal communities.

INTRODUCTION

Most maternal deaths are avoidable, and the solutions to minimize maternal death are well known.¹ Still, in 2016 about 830 women died every day due to maternal health complications related to pregnancy.² About 99% of global maternal deaths occur in developing countries—most often in rural areas, which contributes to over half of these deaths.¹ Authorities worldwide have been trying to minimize maternal deaths by fostering health awareness campaigns at the community level.³ However, the effectiveness of these health interventions or campaigns in rural societies has been negatively impacted primarily by 3 major factors: (a) Inadequate antenatal care delivery, (b) Lack of maternal health awareness, and (c) Tribal belief systems.⁴

Improving awareness regarding antenatal care (ANC) is one of many ways to minimize maternal health risks. ANC encompasses health awareness programs, such as prevention, detection, and treatment of existing ailments concerning maternal and child health.⁵ However, evidence shows that ANC has often been poorly executed, underutilized,^6,7 and disproportionately spread across the socioeconomic gradient of developing nations.^8–10 The lack of accessible healthcare resources^11,12 also restricts people from availing themselves of modern medicines. In addition, tribal and rural communities typically refrain from modern healthcare¹³ and consider maternal death to be normal.^14,15 Lack of maternal health awareness has led to iron deficiencies¹⁶ (anemia and restricted fetal growth), tetanus toxoid (TT) infections,^17,18 and poor hygiene (failure to comply with the “five cleans”: clean hands, surfaces, umbilicus, cord tie, and wrappings for the baby), causing harms, such as genital tract infection and puerperal sepsis.¹⁹

Consumer health technology, particularly mobile health (mHealth) technology, serves as a powerful tool for patient education.²⁰ The Global Observatory for electronic health describes mHealth as healthcare-related practice assisted by mobile devices.²¹ mHealth technologies have substantial potential to facilitate unprecedented and tailored access to treatment advice.²² They are typically used as digital tools to improve healthcare access by minimizing intervention costs and eliminating geographic barriers. mHealth’s text messaging interventions have improved maternal health by guiding women through the various stages of pregnancy.²³ mHealth technologies have also demonstrated promising impacts on several public health issues, including patient management,²⁴ communication in rural areas,²⁵ family planning,²⁶ and diabetes management.²⁷

Mobile health applications are used in various areas; however, there is limited research capturing their effectiveness as an educational tool, particularly in maternal health in developing countries and rural communities.²⁸ Thus far, most research in this domain has emphasized developed nations²⁸ where the consumers (mHealth users) are educated and/or typically familiar with smartphones in general. However, mHealth projects are now being used in developing nations; therefore, it is essential to keep in mind the consumers’ ability to interact with mobile phones. Furthermore, given the challenges faced in medical, especially maternal health, education in developing countries (particularly issues of accessibility and affordability), user-centered mHealth has enormous potential as a tool for teaching and spreading awareness.

This study implements a mobile health technology (mobile for mothers, MFM) as a supporting tool to improve maternal health awareness among tribal communities using a randomized quasi-controlled intervention.

MATERIALS AND METHODS

The study was part of a larger study conducted in collaboration with the Rural Health Mission of the Government of Jharkhand under the European Union-funded Initiative for Transparency and Good Governance. The study received ethical clearance from the Institutional Review Board Center, New Delhi, India. With consideration to the low literacy level among the study population, oral consent rather than written documentation was obtained. The consent form was read in Hindi by one of the project team members. All participants were informed that their participation was voluntary and that they had the right to skip questions they felt uncomfortable with or stop answering at any time. Furthermore, they were informed that the data collected was intended for research purposes only.

Study design

This was a quasi-controlled, cross-sectional analysis of 2 groups (rural villages in India): (a) an intervention group receiving government mandated programs through a mHealth application in Village A, and (b) a standard control group receiving government mandated care programs through traditional means (orally) in Village B. These 2 villages have similar socio-demographics. Under the government programs, community health workers visited pregnant women in both villages and discussed maternal health concerns and measures orally (one-to-one), ensured ambulance availability if needed, and provided financial incentives for women delivering at the hospital. In the intervention village, the community workers leveraged mHealth (MFM) technology to discuss maternal health concerns and measures. All communication occurred in their native language, Hindi.

Mobile health technology/mobile for mothers (MFM)

MFM, a software application, was conceptualized by Network for Enterprise Enhancement and Development Support (NEEDS), an Indian nongovernmental organization (NGO), and Simavi, a Dutch NGO. It was designed for low-literate users to operate on affordable Java-enabled phones or Android-based smartphones that run free and open-source applications containing registration forms, checklists, tracking of danger signs, and instructional prompts. The MFM application consists of 4 modules: (1) Registration, (2) antenatal care, (3) intranatal care, and (4) postnatal care. In addition, the Interactive Voice Recording System enabled mHealth to provide maternal health information through texts, photographs, and voice prompts (in the user’s native language) to pregnant women and mothers, as illustrated in Figure 1. All written information and voice recordings were in the Hindi language .

Figure 1.

Mobile for Mothers application (This is representative, the original language was Hindi).

Intervention

The intervention was led by trained, accredited social health activists (ASHA). ASHA is a community-based health worker program founded as part of the National Rural Health Mission by the Indian Ministry of Health and Family Welfare. ASHA staff visited each participant (mother) 4 times in the prenatal phase (the first visit occurring in the first trimester) and twice during the postnatal phase (third and sixth month after childbirth), as illustrated in Figure 2. The control group also received the same number of visits; however, ASHA workers did not use MFM with the control group. Each village was allocated 400 ASHA workers (800 ASHA workers total).

Figure 2.

Mobile for mothers intervention.

During each home visit in the intervention group, the ASHA carried and used the mHealth application to teach pregnant mothers about maternal healthcare and hygiene, as appropriate, at different stages of pregnancy. The intervention took place with one pregnant mother at a time (one-to-one counseling). Note that in the control group, the same discussion about maternal health and hygiene occurred orally (one-to-one). Each encounter lasted for approximately 45 minutes.

Data collection

Pregnant mothers between the ages of 18 and 45 years were recruited from the control and intervention village, respectively. In January 2014, a team of trained project members administered a paper-based survey to the mothers in 2 villages (control and intervention) to collect baseline data. Team members read the questions and marked the responses for all the participants including both illiterate and literate mothers. Note that ASHAs were not involved in any survey data collection. The survey questionnaire was guided by the National Family Health Survey (NFHS)²⁹ to assess maternal health and hygiene awareness, calculated as a binary variable where women were deemed aware of maternal health information if they responded “yes” correctly to the awareness questions suggested by NFHS (see Supplementary Appendix A).

The mHealth intervention commenced soon after the baseline survey was completed in early February 2014. The end-line (postintervention) data were collected between November 2015 and January 2016. A team of trained project members who were familiar with the region manually collected data by door-to-door visits using the same approach as in baseline. Discussion confidentiality was maintained, and no family members were allowed during the intervention and data collection to prevent external influences on the respondents.

Data analysis

A priori power analysis was completed to estimate the minimum sample size for the study. This analysis included 2-tailed assumptions, an estimated power of 0.80, an alpha error probability of 0.01, and an effect size of 0.2. The results of the a priori power analysis supported the inclusion of at least 1172 participants in data collection.

First, we calculated descriptive statistics related to demographics. Then we calculated the percentage changes for each variable between baseline and end line in each group. Finally, we used the Pearson chi-square test of independence at a 99% confidence interval to compare the percentage changes for each variable between intervention and control groups. All analyses were conducted in SPSS.

RESULTS

The survey consisted of 1480 respondents, 740 women per group. Table 1 shows the demographic characteristics of the respondents.

Table 1.

Demographics of the study population

	Control	Intervention
	N (%)	N (%)
Duration of stay in the village
Less than 5 years	277 (37.4)	302 (40.8)
5–10 years	338 (45.7)	35 (41.2)
11 years and above	125 (16.9)	133 (18.0)
Caste
Scheduled Caste (SC)a	58 (7.8)	100 (13.5)
Scheduled Tribe (ST)b	22 (3.0)	27 (3.9)
Other Backward Castes (OBC)c	523 (70.7)	542 (73.2)
Other than SC/ST OBC	63 (8.5)	69 (9.3)
Education level of women
Illiterate	384 (51.9)	385 (52.0)
Primary (1–5years of schooling)	124 (16.8)	141 (19.1)
Secondary (6–10 years of schooling)	199 (26.9)	187 (25.3)
Higher (11 and above years of schooling)	33 (4.5)	27 (3.6)
Occupational status of women
Working	56 (7.6)	95 (12.8)
Housewife	684 (92.4)	645 (87.2)
Age during intervention
18–19	135 (18.2)	144 (19.5)
20–24	383 (51.8)	368 (49.7)
25–29	160 (21.6)	158 (21.4)
30–34	47 (6.4)	49 (6.6)
35–45	15 (2.0)	21 (2.8)
Age at marriage of women
Below 18 years	471 (63.6)	501 (67.7)
18 years and above	269 (36.4)	239 (32.3)
Religion
Hindu	684 (92.4)	556 (75.1)
Muslim	47 (6.4)	173 (23.4)
Christian	4 (0.5)	8 (1.1)
Sarna	5 (0.7)	3 (0.4)

^aPeople belonging to Scheduled Castes (SC)—otherwise known as “dalits”—are officially protected groups of individuals in the Constitution of India. This sub-community of the Indian caste system faces deprivation, oppression, and social isolation on account of their position at the very bottom of the Indian caste system and perceived low status.³⁰

^bPeople belonging to Scheduled Tribes (ST) are indigenous individuals with primitive traits, distinctive culture, and geographical and social isolation.³⁰

^cPeople who are identified (by the State and Central Government of India) as socially, economically, and educationally disadvantaged. However, there is still no clear definition for OBCs in the Indian Constitution.³¹

Table 2 shows all health and hygiene awareness of the study participants before (baseline) and after (end-line) intervention. The baseline and end-line columns report the percentage of participants who responded correctly to the corresponding questions. The incorrect responses were recorded separately as “No correct knowledge.” The Pearson chi-square compares the change in awareness between the control and intervention groups between baseline and end-line.

Table 2.

Comparing maternal health awareness in intervention and control groups

	Site	Baseline	End-line	Change from baseline	Pearson Chi-square Comparing the percent change (C) between the control and intervention group		Hypotheses
		A	B	C
		(%)	(%)	(%)	χ2	P value
Hypothesis1 (H1): mHealth can improve pregnant mother’s awareness about the need to visit doctors/ANC during pregnancy.
Do you know why you need to visit Doctor/ANC during pregnancy?
Helps to identify (severe) problems and provide solutions	Control	21.0	9.30	−11.70			Accepts H1
	Intervention	9.10	33.10	24.00	125.21	<.001
It will minimize risks	Control	13.3	11.6	−1.7
	Intervention	6.5	19.1	12.6	15.74	<.001
To receive TT and iron tablets	Control	19.5	48.1	28.6
	Intervention	7.7	68.9	61.2	66.01	<.001
No correct knowledge	Control	53.3	31.5	−21.8
	Intervention	80.4	6.2	− 74.2	154.45	<.001
Hypothesis 2 (H2): mHealth can improve pregnant mother’s awareness about the need to take TT injections during pregnancy.
Do you know why you need to take TT injection?
It will protect the baby from tetanus	Control	38.5	49.8	11.3			Accepts H2
	Intervention	17.7	72.7	55.0	96.71	<.001
Hypothesis 3 (H3): mHealth can improve pregnant mother’s awareness about the need to take iron tablets during pregnancy.
Do you know why you need to take iron tablets?
It increases blood and will protect from anemia	Control	33.1	64.4	31.3			Accepts H3
	Intervention	12.0	70.0	58.0	44.55	<.001
Hypothesis 4 (H4): mHealth can improve pregnant mother’s awareness about the danger signs during pregnancy.
Do you know the danger signs during pregnancy?
Bleeding from vagina during pregnancy	Control	21.6	14.8	−6.8			Accepts H4
	Intervention	9.5	27.8	18.3	99.32	<.001
Severe abdominal pain during pregnancy	Control	8.6	10.5	1.9
	Intervention	4.5	27.7	23.2	129.11	<.001
Severe headache with blurring of vision	Control	4.9	24.3	19.4
	Intervention	3.0	23.0	20.0	24.35	<.001
Convulsions/loss of consciousness	Control	3.1	5.5	2.4
	Intervention	1.6	6.5	4.9	10.13	.002
No correct knowledge	Control	71.0	56.1	−14.9
	Intervention	87.4	19.7	− 67.7	21.92	<.001
Hypothesis 5 (H5): mHealth can improve pregnant mother’s awareness about the danger signs during labor.
Do you know the danger signs during labor/delivery?
If the membranes/water is broken/coming out before the expected date	Control	7.5	18.5	11.0			Accepts H5
	Intervention	5.5	40.7	35.2	68.86	<.001
Convulsions/loss of Consciousness	Control	4.2	6.2	2.0
	Intervention	1.6	17.8	16.2	37.13	<.001
Failure to progress	Control	6.7	33.5	26.8
	Intervention	3.1	51.3	48.2	46.27	<.001
The placenta does not come out within 30 minutes after the baby is delivered	Control	2.5	10.8	8.3
	Intervention	1.2	23.3	22.1	34.71	<.001
No correct knowledge	Control	82.5	45.6	−36.9
	Intervention	91.4	16.6	− 74.8	49.57	<.001
Hypothesis 6 (H6): mHealth can improve pregnant mother’s awareness about the importance of the “5 cleans”
Do you know why the 5 cleans are important during delivery?
It limits infection (and death) both in baby and mother	Control	36.1	44.7	8.6			Accepts H6
	Intervention	9.9	67.0	57.1	65.04	<.001

Our analyses indicated significant improvements after the mHealth deployment in the intervention group compared to the control group. Although improvements were also noted in the control group, the magnitude of improvements were significantly more in the intervention group. We also observed negative change in the control group, as marked in red (see Table 2), where awareness about ANC and danger signs of pregnancy was reduced.

Awareness about ANC during pregnancy significantly (P < .001) improved in the intervention group. Participants acknowledging the role of ANC in enhancing health, minimizing risks, and as a source of necessary tetanus toxoid (TT) injections and iron supplements increased by 24%, 12.60%, and 61.20%, respectively. The number of participants who responded incorrectly in the baseline survey also decreased by 74.2%. Contrastingly, awareness about ANC was noted to decrease in the control group, as shown in Table 2. Awareness about TT injections and consumption of iron tablets were also significantly (P < .001) improved among the intervention group by 55% and 58%, respectively. Awareness about hygiene (the “five cleans”) significantly (P < .001) increased by 57.1%, and incorrect knowledge about the same was significantly (P < .001) decreased by 62.40% in the intervention group. Mothers in the intervention group who recognized vaginal bleeding, severe abdominal pain, severe blurring of vision, or convulsions as danger signs during pregnancy significantly (P < .001) increased by 18.30%, 23.2%, 20.0%, and 4.90%, respectively. Similarly, mothers in the intervention group exhibited a higher level of awareness regarding danger signs during labor. Danger signs such as “failure to progress” and “placenta does not come out within 30 minutes after the baby is delivered” were identified by 48.2% and 21.10% more mothers in the intervention group.

DISCUSSION

As maternal healthcare shifts to home and community-based settings, consumer health technology applications hold promise for augmenting patients’ awareness and, in turn, their ability to provide self-care. This study demonstrates how tribal communities—who typically have no to minimal familiarity with mobile devices—learned information about maternal health and hygiene when delivered through an mHealth application in a user-centered manner (using the local language and audiovisual communication). This study also indicates how the mHealth application significantly enhanced the education and maternal health knowledge of pregnant women when added to the typical standard of care in rural and tribal communities.

Our findings showed a significant improvement in the mHealth intervention group’s awareness of critical knowledge during the pregnancy, including the importance of doctor visits, TT injections, iron tablets, danger signs, and hygiene (the five cleans). The findings exhibit the potential of mHealth as an educational and awareness tool that provides structured information about maternal health for pregnant mothers in rural and tribal communities. This study reiterates that health-related education holds value for patients in remote regions to support self-management.²²

Our study highlights the relevance of health literacy to mHealth. Proponents of digital health may acknowledge the dependence of mHealth on overall literacy and how low literacy deters the effectiveness of mHealth interventions in indigenous communities.³² However, our findings show that, if used correctly, mHealth interventions can be an effective audiovisual tool to educate people who have low literacy. The extent to which a user comprehends any information delivered to them by an mHealth application determines the effectiveness of the technology.³³ Despite most participants having no or minimal literacy, the mHealth intervention MFM effectively improved maternal health awareness of pregnant mothers as they easily adopted short and easy-to-read health information. Availability of mHealth content in local tribal languages also contributed to successful mHealth adoption. These findings were in line with other studies,^34–36 which incorporated local language into their software.

In support of existing literature,^37,38 our study demonstrates that mHealth applications have tremendous potential for supplementing traditional channels of maternal health communication (campaigns, posters, public announcements) as an evolving communication medium for fostering maternal health awareness. According to the Cognitive Theory of Multimedia Learning, people learn more effectively from words and images than words alone.³⁹ This theory can partially explain why tribal communities learned new information through mHealth significantly more than traditional health intervention. The use of both auditory and visual channels in mHealth (Dual-Coding Theory) potentially helped pregnant mothers learn new knowledge. However, these assumptions require further exploration and confirmation.

Finally, this study also exhibits the potential of mHealth to minimize anchoring biases of indigenous communities⁴⁰ where their healthcare practices and beliefs are primarily determined by their faith in traditional knowledge, such as natural medicine, psychosomatic treatments, and religious rituals.⁴⁰ Despite having a strong belief in the “traditional health care system,” tribal communities in the intervention group (using mHealth tool) were noted to embrace the scientific or modern maternal healthcare practices in addition to their traditional beliefs.

Some limitations of our study need to be acknowledged. First, being a quasi-experiment where neither women nor ASHAs were randomized, the findings might have been influenced by confounding factors, such as differences between the villages or the ASHA staff. Second, this study did not measure any health outcomes of the pregnant mothers, but only change in health knowledge. Although not demonstrated in the current study, improved health knowledge may be a precursor to improved health outcomes and would be a promising and essential area of future research. Third, ASHA workers were responsible for carrying the mobile device with them during each intervention. Pregnant mothers, being passive users, only used the mobile application in the presence of ASHAs. Further research is needed to capture the direct impact of mHealth on maternal health awareness in tribal communities when actively used by mothers without receiving assistance from trained personnel such as ASHAs. This research would be dependent on reaching the point where there are sufficient smartphone owners in the tribal communities. Last, the study only focused on 2 villages, though the sample size was sufficient. Although more than half of the sample were illiterate, future studies can focus only on the impact of mHealth use on illiterate populations.

CONCLUSION

The results indicate that the mHealth intervention can improve maternal health awareness and knowledge of tribal and rural communities despite low educational status when used as an education tool by community workers. mHealth holds continued promise for maternal health, but implementers and policy makers should additionally address health system and sociocultural factors that play a significant role in the uptake of recommended maternal health practices, especially in rural communities of developing or underdeveloped countries.

FUNDING

The study was funded by Simavi, the Netherlands, under Grant Number 3312005 and Deutsche Welthungerhilfe under Grant Number WHHInd/1287 (to the author, MMC). The content is solely the responsibility of the authors.

AUTHOR CONTRIBUTIONS

MMC conceived and designed the study, participated in data collection, and approved the final version for submission. AC and OA participated in the literature review, graphical illustration, data analysis, wrote the manuscript, and approved the final version for submission.

SUPPLEMENTARY MATERIAL

Supplementary material is available at Journal of the American Medical Informatics Association online.

DATA AVAILABILITY STATEMENT

The anonymized data underlying this article will be shared on reasonable request to MMC.

CONFLICT OF INTEREST STATEMENT

None declared.