Meeting Report

Deep Dive: Developing Mobile Content for Clients

April 26, 2013

Event

The mHealth Working Group hosted a “Deep Dive” meeting on the development of mobile health information — including messages involving text, voice and images — for clients.

Meeting Objectives

  • Share information on various types of mobilehealth messaging in current use
  • Share best practices for the design of mobilehealth content
  • Acquire “hands-on” experience with the best practices for mobile-health messaging

Meeting Date

  • April 26, 2013
  • Luncheon: 12:30 to 1:00 pm, EDT
  • Discussion: 1:00 to 5:00 pm, EDT

Location

The meeting was held at the HolidayInnInnerHarbor in Baltimore, Maryland, USA.

Deep DiveAgenda

12:30 – 1:00 Lunch

1:00 – 1:10 Welcome —Peggy D’Adamo and Adam Slote, USAID

1:10 – 1:20 Context —Stephen Rahaim, Abt Associates

1:20 – 2:20 Brief (8-10 min) “Firecracker” presentations

  • Hesperian — Sarah Shannon
  • HealthPhone — Nand Wadhwani, The Mother and Child Health and Education Trust
  • FHI 360 — Kelly L’Engle
  • Mobile Alliance for Maternal Action (MAMA) — Kirsten Gagnaire
  • Johns Hopkins University Center for Communication Programs (JHU/CCP) —Nandita Kapadia-Kundu,
  • Catholic Relief Services — Marianna Hensley and Carrie Miller

2:20 – 2:30 Descriptions of workshops, logistics and breakout

2:40 – 3:30 First workshop rotation (50 minutes)

  • FHI 360: Kelly L’Engle and Christine Lasway — Role Model Story
  • Mobile Alliance for Maternal Action (MAMA): Kirsten Gagnaire and Brooke Cutler, MAMA and Carolyn Florey, USAID —SMS messages
  • JohnsHopkinsUniversityCenter for Communication Programs (JHU/CCP): James Bon Tempo and Heidi Good Boncana —Storyboarding for mobile photofilms

3:30 – 3:45 Tea/coffee and snacks

3:45 – 4:30 Second workshop rotation (50 minutes)

  • FHI 360: Kelly L’Engle and Christine Lasway —Data collection
  • Mobile Alliance for Maternal Action (MAMA): Kirsten Gagnaire and Brooke Cutler, MAMA and Carolyn Florey, USAID —Integrated Voice Recording (IVR)

4:30 – 5:00 Report back and discussion

Meeting Highlights

Welcome: Peggy D’Adamo andAdam Slote, USAID

We’ve only touched the tip of the iceberg of mobile technology.

Mobile technology is currently focused on SMS, but who knows what we will be using in 5 years? Ms. D’Adamo predictedthe use of some form of phone-tablet hybrid.

The mHealth Working Group comprisesmore than 1350 individuals representing over 450 organizations in 59 countries.

Context: Stephen Rahaim, Abt Associates

USAID’s SHOPS Project = Strengthening Health Outcomes Through the Private Sector.

Mobile phones are the second-most personal form of communication channel (after face-to-face) but they are difficult to use effectively.

For example: The iPhone home screen icons can be seen as behavior-change messages, but how do we know whethercommunication has actually taken place?

Process for successful mHealth development as applied to behavior change:

  1. Design path: Begin with basic behavior-change fundamentals.
  2. Formative research: Do it! Understand your audience/demographic and the issues related toaccess and fluency with technology.
  3. Creative approach: Start with a theory and think outside the box of SMS (a proper messenger—authority, family member, etc.—can add impact when you only have 160 characters).You must entice people to open your message.
  4. Pre-testing: Use the same audience.Never assume you know what the results will be.And donot test yourself.

You must go through all of these steps with each new tool or product.

Designing for impact concerns more than message content.

Keep sharing lessons and success stories.And recycle your old phones.

Firecracker Presentations

Sarah Shannon, Hesperian Health Guides

Hesperian Health Guides develops “trusted and vetted health content” (e.g., the book,Where There is No Doctor). Hesperian is renowned for “writing haiku” — shorthand for the simplification of complex content into empowering, demystified health information that inspires action.

Hesperian launched the HealthWiki in 2012— a freely accessible repository of Hesperian content available to all. One of the great features about the HealthWiki is that is allows for easy to practitioners for cut-and-paste usage, enabling others to create their own localized and adapted health materials, including for use in mHealth.

Currently, there is content in the flexible HealthWiki platform available in 10 languages (with more on the way, as Hesperian content has been translated into over 80 languages). This enables users to easily switch between languages for Content is easy to digest and useful for multilingual messaging (it is currently available in 10 languages), or for working across languages.

The HealthWiki also features an online, easily searchable image library of over 120,000 of Hesperian’s acclaimed images. Users are able to search by key-words to find relevant images for use in developing their own health materials and mHealth applications (apps).

Hesperian recently launched The Safe Pregnancy & Birth pilot app for AndDroid and iPhone, utilizing the content from the HealthWiki and Image library. The app has been downloaded more than 780,000 times from 183 countries.

The app uses open-source code and it is available in English and Spanish (other additional languages hopefully coming soonare in development).

Nand Wadhani, Mother and Child Health and Education Trust

HealthPhone is a mobile video project that started in India. The project targets women living in slums/rural areas who werenot getting relevant health information (e.g., posters werenot reaching them).

About 50% of the mobile phones sold in India during the firstquarter of 2013 were smartphones.

Video is more efficient than SMS for certain topics (A breastfeeding “demo” was provided as an example.)

The project identified a need to invest in the creation of content for mothers. The project formed a partnership with the Ministry of Health in Maharashtra, and organized a video production team in India. (The project continues to be involved in this partnership).

Content is loaded to the phones on a micro SD card, but hundreds of videos are available for viewing and downloading, thanks to partner organizations.

The menu is translated into the local language; it is also numbered and color-coded for low-literacy users.

The project also uses a mother-child tracking system — a personal health worker visits each family,beginning at pregnancy until the child is two years old.

List members can learn more about HealthPhone by watching these videos:
HealthPhone Introduction
TEDx Talk - HealthPhone: The First Mile Now Reachable

Kelly L’Engle, FHI360

Mobile 4 Reproductive Health (m4RH) began as a research pilot project in 2008.

The project is funded by USAID/PROGRESS. The projected originated in Kenya and Tanzania, and is now moving into Rwanda.

The project is entirely an SMS, menu-driven, opt-in communication system. Users receive information about their chosen contraceptive types.A clinic-locator database has also been added.

The SMS systemsis also useful for monitoring and evaluation: Users first encounter a short survey and a notification that includes informed-consent language.The system uses open-ended and closed-ended questions to collect data from individual m4RH users. A 50% response rate can be obtained.

The project recently partnered with JHU·CCP to scale up: about 100,000 people used m4RH during2012 in 127 out of 129 districts in Tanzania.

Abt Associates is evaluating m4RH’s success in Kenya. An evaluation is also underway in Tanzania.

The project also learned the importance of building a roadmap for content deployment so designers know what to do.

Best practices:

  1. Concept testing
  2. Message development
  3. Message approval
  4. Message testing
  5. Utility testing

Role-model stories (narrative/serial dramas) have been effective means of behavior-change communication. These stories were developed in collaboration with local resources, and they are applicable to any health topic.

Kirsten Gagnaire, Mobile Alliance for Maternal Action (MAMA)

More than one billion women in developing countries own or use mobile phones.

MAMA’s goal: Reach 20 million mothers

In Bangladesh, most women are illiterate or have low literacy.These women can talk on the phone, so interactive voice response (IVR) technology is an option, as well as SMS.

The messages and services are directed not only to women/mothers, but also to their husbands and mothers-in-law. MAMA is also training and involving community-health workers (CHWs).

In South Africa, 80% of women are single mothers; many are HIV positive and experience domestic violence.However, 98% are literate and “tech-savvy,” so the addition of more mobile communication options is possible.

The SMS rates in South Africa are very high, so MAMA uses mobile web instead (askmama.mobi).

The project also focuses on the prevention of mother-to-child transmission (PMTCT) of HIV.

The project is currently developing a program in India.

Nandita Kapadia-Kundu, JHU·CCP:

The team used smartphones to launch mobile health clubs that focused on family planning in Uttar Pradesh, India (UPI). About 97% of UPI households have a mobile phone.

The team targeted married men who were not using a family planning method.

Community-health workers held twice-monthly meetings where members viewed new videos— role-model stories, entertaining behavioral films, physician interviews, etc. — in small by phone then participated in a discussion about the family planning content, how to initiate and facilitate couples communication, where to go for services. The CHWs later made follow-up calls to members and provided a hotline number for members to reach a doctor with any immediate related questions.

The videos supported positive behavior change communication viatopics including long-term and short-acting contraception, couples communication, and general health.

Compassion is very important in India, so nonjudgmental messaging is important.

The members were offered small gifts for their wives (perfume, bindis, etc.) to encourage communication about video/discussion topics.

The program included resources for CHWs: an action plan, tracking sheet, discussion guide, and other tools.

Marianna Hensley/Carrie Miller, Catholic Relief Services

ReMIND (Reducing Maternal & Newborn Deaths) pilot project

Many households in India have mobile phones, but men control the phones.

The literacy levels are low, especially among women.

ReMIND worked with government ASHAs (Accredited Social Health Activists) to develop apps for pregnancy (which required 17 months to complete) and the postpartum period (which required 11 months to complete).

Process:

  1. Shifted from closed-ended to open-ended questions to elicit deeper responses.
  2. Employed user-driven counseling.
  3. Separated the health checklist from user counseling.
  4. Used behavior change communication to address families that did not want girls.
  5. Tweaked the menus tweaked for low-literacy users (including the ASHAs).

Lessons:

  1. Test extensively, then scale up.
  2. Focus on teamwork.
  3. Determine “need-to-know”rather than “nice-to-know” user information —Don’t ask too much.
  4. Share and learn from your partners.

Notes: Workshop Rotations

Role-Model Story Workshop

Kelly L’Engle and Christine Lasway of FHI360

  • Two handouts
  • It takes time to develop these stories.
  • Role-model story delivered via SMS
  • This strategy can complement radio stories or other mediated messaging.
  • Tell the story from thefirst-person perspective; the story should be situated in the cultural context.
  • Conduct abehavior-change scan:What does the white and grey literature say about the motivators and the barriers ofthe target behavior?
  • Your messages are limited by the number of characters that can fit on a screen—only 160. Every word should be meaningful and necessary.
  • The stories are made up of installments. An installment is 2 to 4 SMS screens sent at one time. Send installments every couple of days, with a maximum of 8 installments per story.
  • If you raise a concern in an installment, that concern should be addressed in the same installment. Donot leave the reader hanging and do notend on a negative thought because that installment could be the last the reader sees!
  • These stories provide a good platform to demonstrate how a role-model character uses health services.For example, you can show that services areyouth-friendly, employ nice counselors, and don’t require much time.
  • A story can end on a fun note; and the story’s elements can build a relationship with the user. Make sure the story tells a story.
  • The stories are created by experts in behavior-change communication (BCC) and health content.
  • Users must opt-in to receive stories
  • JHU-BCC used a campaign involving posters, TV, radio to increase enrolment.
  • Questions: Can an installment or a frame end in the middle of a sentence or a word?Or should you finish the thought before going to the next screen? How do you know where to cut?
  • We donot have this down to a science. We do make sure to address any negativity within the same installment.
  • We donot use shortcut words (abbreviations) so that everyone can understand themessage. Some projects that use youth-targeted messaging do take shortcuts; and those messages may be written by teens.
  • You donot have to have a high level of “book” literacy to send and receive text messages. We believe that text messages can be considered a different type of literacy.
  • We have expanded this role-model storyapproach as part of the m4RH program.
  • Some research has been conducted on these stories, includingquantitative and qualitative studies, and pre-tests and post-tests in a small trial in Ghana. The research showed that the stories were effective. Users could cite small details demonstrating interest and comprehension and a high level of involvement and processing—all linked to more effective behavior change.
  • We hypothesize that in role-model stories there is a higher level of brain processing required for the long story and the long message compared to a short text message (that is purely factual). The higher level of brain processing could possibly be the reason that longer messages have better retention.
  • The technical partner is Text to Change, which is based in the Netherlands and Uganda.
  • We considered adding a phone number to the text messages for people with questions, but this approach would require someone to be on call 24/7. Anotherpossibilityis an automated message that statesthat all questions will be answered within 24 hours.
  • Approximately 10% of m4RH users “ping” the clinic database.
  • The size for the pilot project in Kenya and Tanzaniawas small: about 6 clinics per country. The clinics consideredthe projectas a value-added service; they are happy to offer it to their patients because it will draw more people to use their services.
  • The recurring costs of implementation include the costs of sending messages; we are still investigating ways to address those costs.

Data Workshop

Kelly L’Engle and Christine Lasway of FHI360

Presentation

  • Mobile devices provide an inexpensive way to collect data.
  • M4RH collected over 25,000 points of data showing district locations of Tanzania users.
  • M4RH has approximately 50% response rates from self-administered surveys sent by SMS. The recent survey response rate was higher than that of previous M4RH surveys, because recent surveys were sent sooner (a few days) after users initiated M4RH.
  • M4RH maintains system logs for information queries from users and survey questions sent to users.
  • Survey questions cover user gender, age range, marital status, promotion point, and change in knowledge, attitude, behavior (KAB).

Q&A

  • Imperfect survey responses are settled by rules for cleaning data, as they would be for a survey in any format.
  • The validity of SMS survey responses is suggested by the high level of congruence between SMS responses and questions conducted by phone interviews.
  • There is a lower response rate for open ended questions, probably because it places greater burden on users.
  • It is important to thank users and confirm their response to every question. (See above point about higher response rate for sooner follow up.)
  • Usage data indicates that approximately 25% ofusers (or phones) are repeatedly querying M4RH for health information.
  • It is not known how phone sharing biases the assignment of queries and responses to specific users.
  • Data can be disaggregated by region. GIS spatial analysis can be added to this. Users’ geographic information can be overlaid with separate data for contraceptive use by area. This would allow comparison of queries about contraceptive options to actual contraceptive use by region.

MAMA: SMS – see presentation

MAMA: Considerations on the use of IVR

Brooke Cutler and Mabinty Koroma

  • Cost factors: recording, distribution
  • Reading literacy, mobile literacy
  • Time —Is voice a better way to convey informationbecausethe medium can convey more than a text message?
  • Know your audience
  • Who are the decisionmakers?
  • Top-down approach
  • Position of authority, even superiority
  • Bottom-up approach
  • Engages the mother on her level
  • Emphasizes what the fathercan do
  • Address cultural beliefs
  • Example: Address the myth that mothers should wait to breastfeed until after the initial period of “darker colored” breast milk
  • Reinforce behavior change

Example of IVR created by the participants in the workshop:

  • You know my wife is pregnant for the first time. I really want my wife and baby to be healthy. What do I need to know?
  • Mom: Do you know if she’s received her shots?
  • Husband: I don’t know. What should she be doing?
  • Mom: Next time your wife goes to the clinic for a check up, remind her to ask about shots she might need shortly after she delivers.
  • Husband: Thanks, mom. Anything else I should know?
  • Mom: Yes.Make sure to ask about the shots your baby will need. He or she will need a number of shots during the first year.

Heidi GoodBoncana and James Bobtempo, JHU/CCP