Feasibility and acceptability of text messaging to support antenatal healthcare in Iraqi pregnant women: a pilot study

Introduction

There are 5.4 billion mobile subscriptions in the developing world – that is 78% of all global subscriptions. Mobile penetration in the developing world has grown to 90.2%, but there is still potential for growth [1]. The introduction of information and communication technologies (ICT) in healthcare, especially the application of mobile technology (m-Health), has already transformed healthcare delivery by making it more accessible and affordable across the developing world where health services are often inadequate, inaccessible, costly, dysfunctional, low in quality and unresponsive to the needs of clients [2, 3]. This health need, in accordance with the World Health Organization (WHO) report, will remain unmet until at least 2050; in this context, through its reach and low cost, m-Health has emerged as a viable solution [4, 5].

There are some online resources for pregnant women available through the Internet, but Iraqi women may lack proficiency not only in English but also in basic computing [6]. Moreover, text is better suited to push technology than the Internet [7].

Iraq’s healthcare system has suffered major setbacks due to decades of war and economic sanctions. Furthermore, access to healthcare has been further affected by recent instability and ongoing strife. Over the past few years, the Ministry of Health (MOH) policy has been to restructure the health system based on primary healthcare (PHC), but the services currently provided at the PHC level only partially meet the health needs of the population. This situation, combined with a low perceived quality of care, has led the population to seek care at secondary and tertiary levels, thereby bypassing primary care [8].

m-Health is an evolving approach that can be defined as “mobile computing, medical sensor and communications technologies for healthcare” [9]. Numerous m-Health pilot projects throughout the developing world have demonstrated key benefits in healthcare outcomes, with implications for healthcare policy [10]. The aim of this project was to introduce an m-Health service to routine antenatal care (ANC) in a PHC setting in a developing country like Iraq, as the mobile phone is a widely (almost universally) used communication device there.

The study objectives were to:

1. Determine the feasibility and acceptability of m-Health technology in antenatal healthcare.

2. Measure satisfaction with text messaging services.

3. Assess the effect of m-Health technology on the count of antenatal visits.

4. Study the association of selected explanatory variables [employment status, educational level, residence type, financial support for the family, healthcare provider, availability of transport facility, ownership of a mobile phone, usage of the Internet to look for health information, current pregnancy status (planned or not) and family history of congenital anomalies] with satisfaction score (positive impact score).

Materials and methods

The project was designed as a controlled experimental study, where an intervention group was exposed to text messages on a weekly basis and the results were compared to those from an unexposed control group. This study was approved by the Ethical and Scientific Committee of Al-Kindy College of Medicine and was undertaken from May 2010 to October 2011. The study was executed in the antenatal clinic that links to Al Elwiya Maternity Teaching Hospital.

A consultant obstetrician designed text messages and supervised the logistics of posting them. The suggested text messages underwent face validity that was assessed by a panel of three obstetricians who spoke the native Arabic language and by 20 pregnant patients as part of a pilot project (unpublished work).

A total of 100 women were recruited for the intervention group and another 150 were recruited for the unexposed (control) group. All the selected females gave their informed consent to participate in the study. Three pregnant women from the intervention group and four from the control group changed their residency place and therefore were lost to follow-up and were, thus, excluded from the statistical analysis. Both groups were selected using a systematic random sample approach in a clinical setting. The average daily workload in an ANC clinic is 60 pregnant females and the working week extends from Sunday to Thursday. The sampling periodicity was set at every 5^th client, allowing for a total of 12 study subjects per working day. Out of each five; the first two were selected in the intervention group and the following three were selected in the control group. This sampling strategy was repeated in the same pattern during the work day. The total sample size was recruited over the course of 5 weeks.

Data were collected from the intervention group using a predesigned structured questionnaire in a personal interview setting.

A weekly text message was sent to each participant in the intervention group starting from the 6^th week of pregnancy. In order to determine the messages sent, client preferences for three types of message themes were explored. These themes included general health messages, those encouraging visits to the PHC center (PHCC) and nutritional advice/lifestyle education. The preference variable was simply a dichotomous one with two choices only for each theme (preferred/not preferred), with the ability to select more than one theme. As a result, the messages sent included 1) reminders to visit the PHCC, 2) general health instructions and 3) dietary advice and lifestyle educational messages. These messages were sent through a website for short message services, “forat-sms.com”, which can generate multiple messages at the same time.

The pre-prepared, validated, Arabic-language messages were sent each Friday between 4 pm and 6 pm as preferred by most of the participants and the sender, as is it is an official holiday. The total cost for sending these messages was $159.885, about 2 cents per message.

The impact of the program was assessed by the degree of satisfaction with its content and process. A total of 11 items were designed to measure the content and process of mobile phone messages. Client satisfaction with these items was explored in the intervention group using a Likert-type scale, which consisted of 5 grades measuring respondents’ agreement with the presented items (ranging from 1 for strongly disagree to a maximum of 5 for strongly agree). The scores on these 11 items were totaled and then presented on a scale with a maximum of 10 grades for clarity and simplicity. For a specific study subject, the score of each of the 11 items was added together, allowing for a maximum score of 55 (11×5) if the subject was highly satisfied with all 11 items (obtaining a score of 5 for “highly satisfied” in each). The resulting total score was weighted to a maximum of 10 by multiplying by a factor of 10/55. The resulting score was called the “total score of positive impact for the intervention program”.

The intervention group participants were also provided with two mobile phone numbers, one for each of the two major mobile phone service providers (Zain Telecom and Asiacell Telecom) to answer any enquiries they had during pregnancy. A total of 314 calls (the average duration for each call was 2.4 min) were received, and confidentiality was ensured.

Both study groups received the same standard healthcare in the clinic. The only advantage of the intervention group was the weekly texts they received and access to the help line.

In addition to collecting demographic information on participants, we calculated the crowding index by dividing household size (count of subjects living in the residence) by count of rooms (excluding the kitchen and bath).

Results

There were no major differences between the intervention and control groups regarding mean age (26.5, SD 6.1 years), median parity (1), employment status (about three quarters unemployed) and gestational age at the time of delivery. The educational level was higher in the control group (Table 1).

Discussion

There are up to 358,000 maternal deaths each year in pregnancy and childbirth. Unlike in the developed world, the risk of maternal death in Iraq is still high; in 2006–2007, the figure was estimated to be 84 per 100,000 by the WHO, adjusted to 74 in 2012 by UNICEF. One of the WHO key working areas is strengthening health systems and promoting interventions focusing on policies and strategies such as m-Health [11, 12]. Moreover, the slow progress and funding constraints in attaining the Millennium Development Goals for child and maternal health encourage harnessing innovative measures, such as m-Health, to address these public health priorities [13]. The issue of competing health system priorities has consistently been rated as the greatest barrier to the adoption of m-Health in various countries [14].

Whilst mobile phone usage is rapidly growing in Iraq, there is still limited use of the Internet, especially among housewives; this was reflected in this study, where only 10.3% of the intervention group reported using the Internet. This makes mobile phone technology an excellent alternative.

This text messaging technology was found to be feasible; the time and effort spent by the provider of the service was reasonable, and it would have been little further work to send messages to an even larger population. It was also shown to be a low-cost method: each message cost $0.0225 at the time of the current study, while a text message ordinarily costs $0.1 (4 times the cost in this text message project).

Based on the satisfaction questionnaire, the frequency of women with high satisfaction (agree/strongly agree with the tested item) was highest (>85%) for “The client recommends this program for other pregnant women”, “Personal rating for the message as a whole”, “They obtained benefit from the messages” and “They would like to be engaged in the same program in their subsequent pregnancies”. It did encourage the participants to visit the PHCC and most participants welcomed the text messages; some wrote positive comments, for example, “Well thank you for your messages that change a lot of wrong information that I have towards my pregnancy”, “I used to finish my work early every Friday waiting for your message”.

Another indicator of the acceptability of this project and its interactive advantages was the 314 phone calls that were received from the intervention group; some women were keen to continue communicating with the service providers even after the deadline of the study was announced (time of delivery).

Although the reminder texts were the least preferred text type, they may have influenced the statistically significant increase in the number (around two visits per pregnancy) of ANC visits, probably encouraging women to attend at certain times in their pregnancy.

This is the first report of using text messaging in the provision of antenatal healthcare of Iraqi pregnant women. Previous published articles have supported the view that a higher count of ANC visits is associated with better outcomes of pregnancy. This view was further tested in low-resource countries; Asundep et al. [15] concluded that <4 ANC visits was associated with an increase in adverse outcomes of pregnancy using adjusted models. Riaz et al. [16] showed that pregnancies with ≥3 ANC visits were associated with far better and healthier outcomes than those with <3 visits, in Pakistan; moreover, the WHO’s latest recommendations that had been released on 7 November 2016 in Geneva stated that a minimum of eight visits for ANC can reduce perinatal deaths by up to 8 per 1000 births when compared to a minimum of four visits [17].

The two comparison groups were homogenous in comparison with respect to known confounders like age, employment status and parity, apart from the possible confounding effect of educational level on the outcome, which was adjusted for by the Poisson regression model. The multivariate model showed that higher education is associated with a lower mean count of ANC visits. This finding complies with the authors’ expectation that women with a higher educational level would experience fewer visits to the PHCC, as they tend to be less satisfied by the quality of services provided at these public centers and generally seek advice at private clinics. This view is shared by Iraqi obstetricians working in the public health services sector. In addition, women with lower education are at greater risk of pregnancy complications, so this m-Health intervention may improve the pregnancy outcomes in this less-privileged population sector.

Our program may offer a new approach to improve ANC services in Iraq. The latest report from the Iraqi Ministry of Health showed underutilization of ANC services by pregnant women, which could be attributed to an absence of awareness about the importance of these services or a lack of confidence in the services available.

The result of this study was comparable with a randomized controlled trial (RCT) conducted in Malaysia using text messaging to improve attendance in primary care: the attendance rate of the text messaging reminder group was significantly higher and the cost of the text messaging reminder was also reasonable, being $0.09 per patient [18]. The same positive result has also been demonstrated in young diabetic people engaged in the Sweet Talk text messaging system trial, where 82% of patients improved their diabetes self-management and 90% wanted to continue receiving messages [19].

Though our text massages had other tasks, their main target was to work as a reminder to attend the PHCC on the scheduled time. In Thailand, automated SMS-based reminders to 280 pregnant women have been shown to increase the number of on-time prenatal care visits by a factor of 1.9 in women who received reminders [20]. In Australia, appointment reminders for attendance at outpatient clinics sent as SMS text messages to patients’ mobile telephones has been shown to significantly lower overall failure-to-attend rates (14.2% vs. 23.4%; P<0.001). Based on this result, the authors plan to extend the trial of SMS reminders to all outpatient clinics in their hospital [21]. Measuring on-time attendance will be one of our targets in a future study.

General health messages, nutritional advice and lifestyle education were the preferred types of messages, which were indicated as the preferred content, exceeding the preference for text reminders. This finding, when viewed in association with the high count of phone calls received by the help center in relation to pregnancy problems, labor information and puerperium, may provide another clue for the acceptability of this m-Health project. The educational role of this electronic method was also shown in other studies. One successful program that is being broadly disseminated in the United States is the “text4baby” program, which sends weekly free text messages to help new and underserved mothers to get access to information that they need to take care of their own health and that of their babies [22].

The explanatory variables taken into consideration in our study failed to show any significant relationship with the positive impact of the messages, apart from family history of congenital anomalies. Future m-Health interventions might therefore be designed to focus on those with a poor obstetric history.

This pilot study showed that mobile-phone-based pregnancy support is feasible, reasonably acceptable by Iraqi pregnant women, of low cost and encourages their ANC visits. Larger randomized studies are needed to provide evidence for extended services with the probable use of an SMS appointment reminder system to improve antenatal visits, to rate the quality of services at the PHC level and, as some women commented that they wished to have more customized communication, to suggest personalized messages tailored to patients’ needs.