Introducing 

Prezi AI.

Your new presentation assistant.

Refine, enhance, and tailor your content, source relevant images, and edit visuals quicker than ever before.

Loading…
Transcript

She takes her baby to a clinic.

At the clinic, Elizabeth hears terrible news.

In 2006, malaria caused nearly one million deaths worldwide.

85% are children under five.

Nairobi has fewer cases of malaria

than many other parts of Kenya.

But Kibera, its largest slum, is home to many newcomers and travelers who may acquire malaria somewhere else.

Kibera

After an infant develops a fever,

she needs malaria medication

within 24 hours.

Only 38% of children in the WHO African region receive any treatment for malaria at all.

Only 3% get the best treatment, ACT.

There are many chemists in Kibera,

East Africa is plagued by stockouts of key drugs.

Unless Elizabeth is able to find

the drug her baby needs today...

this story might not have a happy ending.

Based on the excellent needs-finding reports conducted by our partners at the University of Nairobi

and our assessment that information and access were key pain points,

we set out to create a mobile application that would help Elizabeth and others like her find the medication they so desperately need.

Our initial concept was quite simple: a mobile phone should connect people to medication.

As we began to prototype this idea,

we realized that the system would be a bit more complex.

Our user would need to access a database,

receive information about where to find medication,

and then feed this information back into the database to help the next user.

So we began prototyping each link in this chain.

First, we explored the idea of which medication should be our focus.

Based on feedback we received from experts,

"Look at aretemisinin combined therapies (ACTs - malaria drugs) since malaria is quite common and needs to be treated very rapidly or it can become lethal quickly."

we decided that ACT, the best treatment for Malaria, should be our initial focus.

Next, we prototyped our user interface.

What they said surprised us; price was a much more important consideration for them than we had anticipated.

"I would use this device if it could tell me where my medications were very cheap."

"My friends want to know when your device will be ready."

Our next iteration

reflected this need.

Now, our working prototype of the user interface goes something like this:

Elizabeth would find out about our system at a clinic. This prescription tells her to enter a "shortcode" to start her search.

When she enters the short code,

our system returns a menu of options.

This prototype is now

up and running in USSD.

A major challenge was figuring out how to get the information about drug availability into a database.

We prototyped several possible solutions for this.

We thought about crowd sourcing information about stockouts, but it wasn't clear that enough users would be incentivized to spend money to report the problem.

We considered having pharmacists report their stock, but the needs-finding reports we received from the University of Nairobi showed us that pharmacists might be too over-extended and under-resourced to be asked to do this work.

We explored housing our database at a clinic, but we learned that Tabitha clinic might be the only one with enough resources to do this job, and they already have a well stocked pharmacy in house.

We thought about working directly with a drug manufacturer, and had some initial conversations with the maker of the most successful ACT drug, but we haven't yet established a formal partnership.

Our most promising idea is a partnership with an existing counterfeit protection system.

Since patients already have an incentive to text in, what if they also texted in the location and price information for the drug?

This would solve our incentive problem, and allow us to collect information about location and price that could feed back into our database.

We have spoken with a leader at a social enterprise company doing counterfeit protection work, and we hope to deepen a partnership as our project progresses.

So, that's a bit about how we started here...

went through many rounds of prototyping...

Next Steps

  • test user interface for ease of use and language

  • continue to prototype location descriptions

  • prototype clinic/patient connection and advertising

  • seed database with information and

begin initial machine learning tests

We hope that, someday soon, a baby's fever will not have to be a death sentence.

91% of deaths from

occur in Africa.

malaria

(WHO statistics)

but some do not have the drug Elizabeth needs.

Data reported via SMS to StopStockouts.org

Nairobi also suffers from stockouts.

Elizabeth's

baby has a fever.

We set out to

change this story.

Kifaa cha Tembe

[Pill Check]

database

ACT

for Malaria

- Michaela Kerrissey, Researcher at Harvard School of Public Health who has done mHealth projects in East Africa

We went through several iterations...

Crowd Sourcing -

Patient Reporting Stock-outs

Stop Stockouts week of action using ushahidi.org

Before we arrived at this one, which we showed to users in Kenya.

Pharmacist Sourcing -

Pharmacists Report Stock to Clinics

Salome Muwanga . Abaluhya woman from Kibegari.

Clinic Managed System -

Clinics house and maintain database of drug availability at local pharmacies

If she presses "1"

she'll get back a list of pharmacies that have the drug in stock, as well as their prices.

and receive more information about how to get there.

She can choose any pharmacy by number from the list...

Supply Chain Transparency - Patients get access to information about recent drug shipments

When a patient receives a new drug, she scratches to expose a code. She texts the code and receives an "ok" or "no" to tell whether the drug is real.

database

and ended up here.

Soon, we should be able to use mobile phones to connect people like Elizabeth to the medication they need.

Kifaa cha Tembe

[Pill Check]

Learn more about creating dynamic, engaging presentations with Prezi