Medical tech designed to meet Africas needs Soyapi Mumba

Like every passionate
software engineer out there,

I closely follow technology
companies in Silicon Valley,

pretty much the same way soccer fans
follow their teams in Europe.

I read articles on tech blogs

and listen to podcasts on my phone.

But after I finish the article,

lock my phone and unplug the headphones,

I’m back in sub-Saharan Africa,

where the landscape is not quite the same.

We have long and frequent power outages,

low penetration of computers,

slow internet connections

and a lot of patients
visiting understaffed hospitals.

Since the HIV epidemic,

hospitals have been struggling
to manage regular HIV treatment records

for increasing volumes of patients.

For such environments,

importing technology systems
developed elsewhere has not worked,

but in 2006, I joined Baobab Health,

a team that uses locally based engineers

to develop suitable interventions

that are addressing
health care challenges in Malawi.

We designed an electronic
health record system

that is used by health care workers
while seeing patients.

And in the process we realized that we
not only had to design the software,

we had to implement
the infrastructure as well.

We don’t have enough medical staff

to comprehensively examine every patient,

so we embedded clinical guidelines
within the software

to guide nurses and clerks

who assist with handling
some of the workload.

Everyone has a birthday,

but not everyone knows their birthday,

so we wrote algorithms
to handle estimated birthdates

as complete dates.

How do we follow up
patients living in slums

with no street and house numbers?

We used landmarks to approximate
their physical addresses.

Malawi had no IDs
to uniquely identify patients,

so we had to implement unique patient IDs

to link patient records across clinics.

The IDs are printed as barcodes

on labels that are stuck
on personal health booklets

kept by each patient.

With this barcoded ID,

a simple scan with a barcode reader

quickly pulls up the patient’s records.

No need to rewrite their personal details

on paper registers at every visit.

And suddenly, queues became shorter.

This meant patients, typically mothers
with little children on their backs,

had to spend less time
waiting to be assisted.

And if they lose their booklets,

their records can still be pulled
by searching with their names.

Now, the way we pronounce
and spell names varies tremendously.

We freely mix R’s and L’s,

English and vernacular
versions of their names.

Even soundex,

a standard method for grouping words
by how similar they sound,

was not good enough.

So we had to modify it

to help us link and match
existing records.

Before the iPhone,

software engineers
developed for personal computers,

but from our experience,

we knew our power system
is not reliable enough

for personal computers.

So we repurposed touch screen
point-of-sale terminals

that are meant for retail shops

to become clinical workstations.

At the time, we imported
internet appliances called i-Openers

that were manufactured
during the dot-com era

by a failed US company.

We modified their screens

to add touch sensors

and their power system
to run from rechargeable batteries.

When we started, we didn’t find
a reliable network to transmit data,

especially from rural hospitals.

So we built our own towers,

created a wireless network

and linked clinics in Lilongwe,

Malawi’s capital.

(Applause)

With a team of engineers

working within a hospital campus,

we observed health care workers
use the system

and iteratively
build an information system

that is now managing HIV records

in all major public hospitals in Malawi.

These are hospitals serving
over 2,000 HIV patients, each clinic.

Now, health care workers
who used to spend days

to tally and prepare quarterly reports

are producing the same reports
within minutes,

and health care experts
from all over the world

are now coming to Malawi
to learn how we did it.

(Applause)

It is inspiring and fun

to follow technology trends
across the globe,

but to make them work

in low-resourced environments

like public hospitals
in sub-Saharan Africa,

we have had to become jacks-of-all-trades

and build whole systems,
including the infrastructure,

from the ground up.

Thank you.

(Applause)

像所有热情的
软件工程师一样,

我密切关注
硅谷的科技公司,

就像足球迷
关注他们在欧洲的球队一样。

我阅读科技博客上的文章,

并在手机上收听播客。

但写完文章,

锁上手机,拔掉耳机,

我又回到了撒哈拉以南非洲,

那里的风景不太一样。

我们长期频繁停电,

电脑普及率低,

互联网连接速度慢

,很多病人
到人手不足的医院就诊。

自 HIV 流行以来,

医院一直在
努力管理定期 HIV 治疗记录,

以应对越来越多的患者。

对于这样的环境,

导入
其他地方开发的技术系统并没有奏效,

但在 2006 年,我加入了 Baobab Health,

该团队使用当地

工程师开发合适的干预措施

,以应对
马拉维的医疗保健挑战。

我们设计了一个电子
健康记录系统

,供医护人员
在看病时使用。

在这个过程中,我们意识到我们
不仅要设计软件,

还要
实施基础设施。

我们没有足够的医务人员

来全面检查每位患者,

因此我们在软件中嵌入了临床指南

以指导

协助处理
部分工作量的护士和文员。

每个人都有生日,

但不是每个人都知道他们的生日,

所以我们编写了算法
来处理估计的生日

作为完整的日期。

我们如何跟进
生活在

没有街道和门牌号码的贫民窟的病人?

我们使用地标来近似
它们的物理地址。

马拉维没有
唯一标识患者的 ID,

因此我们必须实施唯一的患者 ID

来链接诊所之间的患者记录。

ID 以条形码的形式打印

在标签上,这些标签贴

每位患者保存的个人健康手册上。

使用此条码 ID,

使用条码阅读器进行简单扫描即可

快速调出患者记录。

无需

在每次访问时在纸质登记簿上重写他们的个人详细信息。

突然,队列变得更短了。

这意味着患者,通常是
背着小孩的母亲,

必须花更少的时间
等待获得帮助。

如果他们丢失了小册子,

仍然可以
通过搜索他们的名字来提取他们的记录。

现在,我们发音
和拼写名字的方式千差万别。

我们可以自由地混合使用 R 和 L、

他们名字的英文和白话版本。

即使是 soundex,

一种根据发音相似程度对单词进行分组的标准方法

也不够好。

所以我们不得不修改它

来帮助我们链接和匹配
现有的记录。

在 iPhone 之前,

软件工程师
为个人电脑开发,

但根据我们的经验,

我们知道我们的电源系统

对于个人电脑来说不够可靠。

因此,我们将原本用于零售店的触摸
屏销售点终端改造

成临床工作站。

当时,我们进口了
名为 i-Openers 的互联网设备

,这些设备是
在互联网时代

由一家失败的美国公司制造的。

我们修改了他们的屏幕

,添加了触摸传感器,

并且他们的电源
系统由可充电电池供电。

刚开始时,我们没有
找到可靠的网络来传输数据,

尤其是来自农村医院的数据。

因此,我们在马拉维首都利隆圭建造了自己的塔楼,

创建了无线网络

并连接了诊所

(掌声)

医院校园内工作的一组工程师,

我们观察到医护人员
使用该系统

并反复
构建一个信息系统

,该系统现在正在管理

马拉维所有主要公立医院的 HIV 记录。

这些医院为
每个诊所的 2,000 多名 HIV 患者提供服务。

现在,
过去需要花费数天时间

来统计和准备季度报告的卫生保健工作者

正在几分钟内生成相同的报告

来自世界各地

的卫生保健专家现在来到
马拉维了解我们是如何做到的。

(掌声

)关注全球的技术趋势是鼓舞人心和有趣

的,

但要让它们

在撒哈拉以南非洲的公立医院等资源匮乏的环境中工作

我们必须成为多面手

,建立整体 系统,
包括基础设施,

从头开始。

谢谢你。

(掌声)