In 2013, I joined Jhpiego to implement their pilot mobile health (mHealth) project in Indonesia. The subject was relatively new in the country, and to my knowledge, there were only three organisations working in mHealth at the time. Despite this, I was the confident that we would successfully roll-out the project. Here is how we planned and implemented the project.

The mHealth project aimed to provide free educational and health messaging service, by means SMS, for expecting and newly-delivered mothers living in rural areas.  We particularly wanted to reach out women who were disconnected from health system and at the same time, encouraged them to visit community health service on regular basis during the period of pregnancy and after delivery. We spent the first 6 months for preparation and planning, from setting up the SMS system to negotiating bulk SMS price to marketing strategies, and on the next 8 months to pilot the service in two districts, each with more than 2 million population.

Choosing a catchy name

The project proposal, a rather short – two pages document, explained background, needs and approach. The project title was “Leveraging Mobile Technology to Save Lives in Indonesia”. Whilst looking good on proposal, we thought the project name was a bit long and inconvenient. We had to find new name for our SMS service so that our target groups: government officials, Community Health Workers (CMW) and mothers can remember.

We came out with the idea of SMSBunda, bunda means mothers in Indonesian. Our country director agreed with the new name. The staff Ministry of Health (MoH) followed the step. From now on, everyone simply refer the project as SMSBunda.

Developing health related contents

The example of SMSBunda content. Photo: EMAS Indonesia

Content is the crucial part of the SMS service. We focused on delivering health content from first trimester until 42 days after delivery. If necessary, additional content could be added on the following phase. We translated and adopted some messages from Mobile Alliance for Maternal Action (MAMA) that was already used in other developing countries. We also worked with health experts from Jhpiego, our organisations partners and MoH to create new contents suitable for Indonesia.The process took around 3-4 months include negotiation and consultation with key stakeholders as well as approval from MoH. The overall characteristics of the content that were developed are:

  • Inline with the guidance of MoH
  • Mainly aimed for mothers but fathers should be able to understand the message
  • Covers antenatal and postnatal care such as nutrition, health myths and facts, emergency and family planning
  • Suitable with local values, for example, content related with alcohol was opted out assuming majority of our target groups were moslems
  • Each content is up to 140 characters. Content beyond this limit should be divided into two text messages.

Setting up SMS system

We had two options in regard with SMS system development. First option was building our own system from the scratch. This was less favourable due to resource time constraint and staff shortage. The second option was adopting an existing SMS system that had been used in real setting environment. We looked at various SMS systems available in the market such as TextIt, CommCare, Vumi and Telerivet as well as SIJARIEMAS. SIJARIEMAS is a referral exchange system developed and used by our partner. We analysed and tested these systems, and we decided to use Telerivet which fits nicely with our requirements as follows:

  • Automatic subscription through SMS. Mothers should also be able to unsubscribe from the service whenever they want. Alternatively, users data can entered manually from web-based interface.
  • To schedule message delivery based mother’s estimation date of delivery, each register mother can receive up to 3 text messages per day.
  • To send out high numbers of text messages, real time and without interruption.
  • Installation and maintenance cost.
  • API connection with mobile network operators and short code.
  • To generate statistical reports.
  • Other functionalities such as SMS survey.
  • Ownership

Stakeholders buy-in and promotion

Photo group: my colleagues and I (middle-right) after project launch event. Photo: United Communications

It should be noted that SMSBunda project did not aim to replace the roles of CMW. In contrary, it complemented the existing health interventions to provide information for mothers.

Therefore, we directly worked with key stakeholders from national and local government. We trained CMW to register their clients and as much as possible linked our activities with theirs.

With the support of an advertising agency in Jakarta, we promoted the service through health facilities, local radio and television channels, posters, billboards, and national and local newspaper, online and offline.

By mid of 2014, 6 months after the project officially launched, more than 30,000 SMS were delivered to 2,500 mothers.

But there were some obstacles too

Obtaining local short number was more complicated than expected, and because of the government policy, we were unable to use short code offered by SMS aggregator companies abroad. Another consequences using short code was premium charge imposed to SMS sender (i.e. project implementer) and receiver (i.e. mothers).

We connected SMS system with Android phone and Telkomsel API – the biggest mobile operator in the country. These combination allowed non Telkomsel customers to benefits from our service. The drawback was time discrepancy of SMS sent using API connection and Android.

The SMS system was designed for one-way communication. However, mothers began to reply the messages with more questions regarding their pregnancy. They had no knowledge that our system automatically generated the message. We answered their questions but soon overloaded with this new task.

The cost to send one SMS was relatively cheap, around 180 IDR (1 cent). But as we reached out more mothers, the SMS cost increased dramatically. During the pilot, we could not predict how many mothers would register to the service and how many of SMS would be delivered. The number of pregnant women and newly-delivered mothers in Indonesia in 2014 was over 10 million, 10% of them resided in our pilot districts.

Some of the registers mothers unsubscribe from the service. They might not find the content useful or switch mobile SIM card. Before I left the organisation, I helped designing M&E activities to improve the service. From what I heard, they used my inputs for the study.

Note:

Jhpiego has expanded the service to many parts of Indonesia. For more details, please visit their website

 

No more articles