Mhealth and Utilization of Health Care Services in Lagos Southwest Nigeria- A Pilot Study

Background: A third of pregnant women in Nigeria do not attend Antenatal care, hence this quasi-experimental study was carried out within 500 consenting pregnant women within the 5 administrative zones of Lagos to determine the association between Mhealth services and utilization of antenatal and skilled delivery services amongst pregnant women in Lagos, Southwest Nigeria. Methods: Participants were divided into an intervention (who received voice messages) and a control arm (did not receive voice messages). Ownership of a mobile phone and willingness to participate were the selection criteria. Data was collected using a structured interviewer administered questionnaire. Analysis was by descriptive statistics with 95% CI to identify factors associated with antenatal attendance. Results: Total respondents were 488 (response rate 97.6%). One hundred and fty-nine (63.8%) of the intervention group found the messages useful. Using the WHO 2016 ANC model, women in the intervention group had a signicantly higher frequency of antenatal care attendance than the control group (p < 0.0001). There was a statistically signicant difference in antenatal attendance between the intervention and the controls (p < 0.0001). There was also a signicantly lower likelihood of poorly supervised (unskilled) home deliveries within the intervention arm (p=0.011). Conclusion: Mhealth was associated with an increased antenatal attendance and skilled birth attendance at delivery. This signicant in reducing and mortality


Introduction
Nigeria accounts for 20% of the global maternal death rates with a maternal mortality ratio of 917 per 100,000 live births as at 2017. 1 In Nigeria, 56.8% of pregnant women are said to attend a minimum of 4 Antenatal care visits at any facility during the course of their pregnancy and 67% receive this Antenatal care service from a skilled birth attendant. 2 There is increasing evidence that this unremitting high maternal mortality rate is partly due to poor quality antenatal care, lack of access to antenatal care as well as underutilization of quality care amongst other causes . 3-5, Several programs and strategies have been put in place by the government to mitigate against these numerous causes, in spite of this only 39% of births are said to be attended to in a health care facility.2 Su cient evidence has shown that effective communication and engagement with health care providers is commensurate with uptake of health care services and simple text-based interventions can help improve utilization of antenatal care services. 4,5 Mobile Health technology (mHealth) includes the use of devices such as the mobile phones, smart phones, handheld and ultraportable devices such as Laptops and Ipad to mention but a few. These devices have become more appealing to health care providers because of their popularity, mobility and technological pro ciencies as well as their ability to deliver health information and services to remote locations. [6][7][8][9] These devices can be used to report health information and deliver health care services through telemedicine in developing countries and as such help improve patient education and adherence to appointments, disease self-management and remote monitoring of patients. 6 Evidence has also shown that a simple device such as a mobile phone minimizes the time constraints encountered in accessing healthcare and facilitates urgent referrals for emergency obstetric and maternal health services in low and middle-income countries (LMIC). 10 Today, Mobile phones have reached the hands of 90% of the world's population, 80% of whom dwell in the rural areas. 7 This rapid development of mobile technology has generated new ways to address public health challenges and raised the standard of health care access and delivery. 8 There have been a few systematic reviews evaluating the role mHealth plays in improving the health care system especially in developing countries. Murthi et al 11 concluded that mobile voice messages provided suitable and appropriate information throughout pregnancy and helped improve uptake of maternal services geared towards improving maternal health outcomes. Free et al7 too found advantages related with mHealth interventions as it concerns utilization of health services especially in the area of appointment reminders and provision of related clinical support. Finally, Piette et al also established useful bene ts in patient self-care and chronic disease management in LMIC through the use of mHealth and Buong et al discovered that short messaging service (SMS) had greater impact than pamphlets in improving the knowledge, attitude and health seeking behaviors of mothers in LMIC. 12,13 Uptake of post natal care in LMIC especially in rural areas is described as very poor, mhealth strategies used to encourage PNC visits was characterized by an improvement in uptake of PNC services. 14,15,16 With emerging interest in this eld, numerous large scale research studies are being done in order to generate the evidence needed to encourage investing in major pilot studies. This study aimed to investigate whether providing care and appointment reminders through mobile phones has any relationship with utilization of maternal health care services such as antenatal care, facility delivery and postnatal care for pregnant mothers and their newborn.

Methodology
This quasi experimental study was carried out in Lagos, a state in the southwest geopolitical zone of A multi-stage simple random sampling technique was adopted for this study to achieve a good representation of the population. All the primary health care centers in Lagos were strati ed into the 5 administrative Districts. Two comprehensive PHC per district were selected by simple random sampling from the list of all the 20 agship primary health care centers in the 5 districts. Using a balloting technique, one of the PHCs was selected for recruitment of the control arm whilst the second PHC was selected for recruitment of the intervention arm as such, there were 5 PHCs as intervention arm and 5 PHCs for the control.
The sample size was calculated using the formula for comparing 2 proportions with a standard deviation at the 95% con dence interval (1.96). The minimum sample size calculated was 227 per group of pregnant women with a non-response rate of 10% this brought the total sample size to 250 consenting pregnant women recruited per group. This study was conducted between April 2019 to September 2019 and all participants were followed up till delivery and two weeks postpartum.
Participant selection criteria included pregnant women at their booking visits in the PHC, ownership of a mobile phone and willingness to participate. The participants were identi ed by a eld worker-based enrollment system and interviewer administered questionnaires were used to gather preliminary data.
Field workers explained the study to all eligible participants by reading the participant information sheets in the preferred language of the participant; a written informed voluntary consent was thereafter obtained and baseline data collection was done.
The participants in the intervention arm received a mobile-phone-based voice message fortnightly consisting of information about the advantage of maternal health care service usage (i.e. ANC, PNC and institutional delivery) as well as appointment reminders. After birth, post natal care visit reminders and information about its advantages were also sent.
Two weeks post-delivery, participants were again requested to ll a questionnaire at their post-partum visit. Information on outcome of pregnancy, place of delivery and client satisfaction of services offered was collected. Clients who defaulted in clinic attendance were followed up via direct phone calls from the eld workers.

Results
A total of 488 women participated in this study (249 intervention arm and 239 controls). Table 1 shows the sociodemographic characteristics of the observed population. The mean age of participants was 28.3 ± 5.5 years. Just over half of all the participants (58.2%) had secondary education. More women in the intervention arm had tertiary education and above when compared with the control group. Majority of the participants were of parity 0-4.  Sixty-two (24.9%) of the participants who were enrolled in the intervention arm failed to receive voice calls for various reasons. Some had provided phone numbers that belonged to their spouses, some missed their voice calls due to lack of battery power on their phones and some said the timing of the calls was too early (7-8am in the morning. A primary health care center selected as one of the controls was found to have a similar service already being provided for their clients (ie reminder text messages) after the study had already started and as such 50 (20.9%) of the respondents in the control arm were also receiving appointment reminders.
Total number of ANC visits are summarized in Fig. 1. This summary is based on the WHO recommendation of at least 8 visits between a pregnant woman and her care giver in the course of pregnancy. The frequency of ANC visits amongst the intervention group was signi cantly higher when compared with the control group using this criteria (p < 0.000).
Of the fteen women who delivered under unskilled care at home, fewer were in the intervention arm than in the control arm (26.7% vs 73.3%) and there was a signi cant difference in the place of delivery in both groups, favoring a higher frequency of institutional delivery in the intervention group (p = 0.011) ( Table 2).  x 2 = chi-square; p = p value The relationship between age, parity, level of education and antenatal attendance was observed in Table   4. There was no statistically signi cant difference noted between most sociodemographic data collected and frequency of antenatal visits for both the intervention and control arm. However, ANC attendance within the control arm was observed to be signi cantly lower in the participants with a lower level of education.

Discussion
This study demonstrated that antenatal care reminders through voice messaging improved antenatal attendance and institutional delivery. Frequency of antenatal attendance was statistically signi cant as well as a reduction in home birth thus strengthening the call for skilled birth attendance in delivery especially in developing countries.
An international survey conducted on 2000 health care professionals, patients and consumers in 4 different countries (USA, China, Brazil and India) on mHealth and healthcare service delivery reported an improvement in health care outcomes with an improvement in quality of care. Patients and consumers interviewed also reported an achievement of their health goals through improved health seeking behaviors and increased compliance with medications and follow up visits. All respondents agreed that uptake of mHealth could be achieved if it was simple, affordable and accessible. 15,16 This study attempted to ful ll these assumptions and observed an improvement in antenatal attendance.
Many studies and systematic reviews on the role of Mhealth and utilization of maternal antenatal and postnatal studies in developing countries have been carried out with similar and promising ndings as observed in this study A detailed analysis of all the studies reviewed established a signi cant increase in maternal and neonatal service utilization with increase in both ANC and PNC attendance. [17][18][19] Shiferaw et al in their study concluded that the educational messages and hospital visit reminders received caused the women to feel more valued by their health care providers and hence made them more responsive to the care and attention they received during pregnancy. 20 All these studies were based on a minimum ANC attendance of 4 visits. This study based its ndings on a minimum of 8 ANC visits per arm and was still able to demonstrate a signi cant increase in attendance. WHO currently recommends a minimum of 8 contacts in pregnancy: ve in the third trimester (fortnightly from 30 weeks), one in the rst trimester (Preferably within the 1st 12 weeks) and two contacts in the second trimester (between 16-20 weeks and 20-26weeks). 21,22 In Sondaal et al review, the major mhealth application was via short messaging service (SMS); however irrespective of the pathway involved, uptake of mHealth was more enhanced if the message being conveyed was in lay terms and in local/preferred language of the patients. 17 This study used mobile phone based calls in interacting with the participants in their preferred language of choice for ease of communication irrespective of the literacy level. Murthi et al conducted their study using the "mMitra" voice messaging service in India and encouraged the use of mobile voice messaging service as it provides a positive impact on maternal health care and improves health outcomes especially in low literacy settings. 22 In Nigeria, a similar study was conducted with pregnant women in Ondo state, participants were provided with mobile phones to see if there was an improvement in utilization of primary health care facility utilization, a 43.4% increase in utilization was seen. 23 In Northern Nigeria, women without mobile phones had signi cantly lower access to ANC services and skilled health care delivery. 24 . No signi cant differences were observed in uptake of post natal services in both studies as was also observed in this study.
Another study on mhealth and post natal care utilization in Nigeria had 63% of its study participants not utilizing post natal care. Promoting education of these women, appointment reminders and mobile clinics were suggested strategies to improve these numbers 27,28 This study attempted to encourage post natal visit by sending appointment reminders 2 weeks postpartum but did not observe any signi cant difference in PNC attendance. WHO recommends at least three postnatal contacts for all mothers and newborns, on day 3 (48-72 hours), between days 7-14 after birth, and six weeks after birth. 29 The participants who did not attend PNC stated that the PNC clinic visit was too soon after birth.

STRENGTH AND LIMITATIONS
Major limitations to the study was the difference in educational status of the two groups. The intervention group were older and more educated than the control group. A randomized trial might have showed a less biased difference. Another limitation was our inability to provide remote monitoring via a two way communication due to cost of software. Calls were made to participants only in cases of defaults to clinic visits to nd out the reasons why. Automated voice messaging systems as was used in this study is a veritable tool for providing Mhealth services especially in developing countries but this requires 3G network with its added cost implication. 30 Also worthy of mention was a part of the control arm who also got text reminders but the fact that despite that, intervention still appeared to work suggests that mHealth was a truly useful tool.
Mobile phones are becoming more inexpensive by the day and as such integrating it into the primary health care setting in resource poor communities is worth considering. In addition, developing alternate power sourced health applications should also be considered in developing countries as most of these countries lack continuous power supply giving rise to inability to fully charge phone batteries. 30 In this study there were cases of study participants not receiving voice calls due to inability to charge their phone batteries following a lack of power supply.

Conclusion
We have found in this study that Mhealth is a useful tool in health promotion and adherence as well as in cultivating a quality health care system however further studies are needed to design an appropriate Mhealth initiative which can be incorporated into the comprehensive health care delivery system. More effort must also be geared towards promoting the importance and uptake of Post natal care.

Declarations
Ethical Approval and Consent to Participate.
The study protocol was approved by the institutional review board (IRB) of the College of Medicine of the University of Lagos Health Research and Ethics Committee (CMUL/HREC/03/19/508) from April 2019 to