07 December 2011

#mhs11 Day 2: The Unknown #FailFaire

The following post originally appears in UN Dispatch.  Two sections were not included in the UND post and I should note are my personal reflections on the conference.  Hopefully, they are more constructive than griping. I would love to hear back from other attendees on how to strengthen the conference next year.
Day 2 at the mHealth summit can be summed up by one word: failure. A session on maternal and child health in Africa offers opportunities in which participants might learn from past mistakes.
Intervention Failure
Dr. Koku Awoonor-Williams discussed the Mobile Technology for Community Health (MoTeCH) project in Ghana; a collaboration between the Ghana Health Service, Columbia University and the Grameen Foundation, he told a story that was also shared at the World Bank’s FailFaire.  The preparation for the project seemingly had all of the elements needed for a successful pilot.  A survey found that 80% of households had access to a mobile phone, making it ideal for implementing a mobile-based health intervention.
Using a snazzy randomized control trial, the program sought to determine if mobile-phone-based health information can improve outcomes and information collection.  The trial was to focus on pregnant and new parents by providing weekly messages in the language and format (sms vs voice) chosen by the client.  Health workers in the trial used the phones to collect patient data.
The trial came back with surprising results about the clients and the health workers.  Despite the survey saying that the majority of women had access to mobile phones, the truth was that the men largely had control meaning that women were not seeing the messages.  The health workers were a bit more mixed.  They liked the simplification of the process, but problems arose when they did not complete all parts of the reports.
What the trial captured was a failure that will lead to an improved intervention; though it does highlight one of the challenges on the ground when it comes to ICTs.  There are plenty of big statistics about mobile phone use and penetration in Africa, but those numbers do not capture levels of access among users.
Study Design Failure
In the same panel, head of D-Tree International Marc Mitchell spoke of a trial to determine of a mobile phone based Integrated Management of Childhood Illness (IMCI) was better than the traditional paper based.  The trail split 18 health centers in Tanzania into a control and treatment group.  The control group was to continue following the standard paper-based IMCI while the treatment group were given an hour training in using mobile phone-based IMCI application.  Mitchell shared results that pointed towards the mobile phone intervention being used more correctly by health workers and improved outcomes for patients.
What this proves is that the treatment group had improved outcomes, but the cause is not quite clear.  One person asked Mitchell if it may have been the additional training that led to an improved understanding of implementing IMCI.  Could the phone be the incentive to act?  Mitchell deferred on the question saying that he would talk with the individual on the side, so the rest of the attendees were unable to know the answer to the question.
The interaction drove home one of shortcomings of mHealth research.  Even when there is impeccable design, the research is often limited to small groups of people in a single region.  The results make it hard to understand how this can be translated to national plans in a given country.
It is where interaction with practitioners is necessary.  These studies do not capture outside influences like when a government strikes a deal with a telecoms company and are forced to use a single network.  They do not take into account when an NGO is donated hundreds of phones and then must implement mobile programs on the donated device.  Such variances are not going to be captured in a small well-executed study.
Failure to Inspire
In day 2, attendees were provided a free lunch.  The price was to sit through a series of keynotes that made for little opportunity to connect with the very people sharing the table.  Each of the speakers were pleasant enough, but did little to captivate attention as people clanked away while eating their salads and roasted chicken breast.  Servers waited in posts to be as helpful as possible by serving salad dressing and trying not to make too much noise as they cleaned off the plates from the first course.  To add a pinch of irony to the lunch, Eric Dishman of Intel managed to frame the idea of moving 50% of health services out of ‘brick and mortar buildings’ in 10 years by saying that mHealth solutions should be practical and avoid hype.

Failure to Engage
I wrote on this yesterday, but it deserves further expanding; the setup of the conference has done little to enable collaborative and engaging discussions.  The problem is not the conference itself, rather it is the format that is utilized.  Each of the panels are packed with presentations that allow enough time for one or two questions per speaker.  Even if Mitchell was willing to engage in the question of trail design, he would not have had enough time to adequately answer the question.  As the sessions end people rush towards the speaker of choice to get in a comment or trade business cards.  Couldn’t this be achieved by reading the published paper and emailing the author?  With over 3,000 people who are involved in mHealth in the same building, the aim should be to find ways to share ideas that can improve health outcomes, program implementation, yada yada yada.

The most apparent disparity can be seen between the international and domestic cohort.  The domestic side seems to hold the all of the cards as most panels are focused on US mHealth.  Even the keynotes lack an international perspective.  There are certainly points of intersection between international and domestic, but it is safe to say that the Tanzanian Ministry of Health is probably not over at the Verizon display to see what they have to offer.

Failure to Fail
This round up should end on a high note, because this has been a great conference so far.  The summit has pulled together a diverse group that will support the growth of mHealth.  As I write this, I am sitting in on a panel that looks critically gathering evidence to know more about mHealth.  The growth of the conference in three years from 300 to 3,000 is a sign of the growing sector.  Innovations that are here in the National Harbor and others designed in the field right now will help to close the gap in health services around the world.