Improving humanitarian response through better health data

Improving humanitarian response through better health data: analysis of two health information
systems in MSF
*Carme Baraldés1 , Jean-François Saint-Sauvuer1 , Inma González 1 ; **Olivier Cheminat2 , Audrey
Landmann2 , Cyril Bousquet 2
1
Médecins Sans Frontières (MSF), Barcelona, Spain; 2 MSF, Paris, France
*[email protected]
**[email protected]
Introduction
MSF needs reliable, complete, and timely information from its medical programmes at project level
and in a format enabling analysis for operational decision-making. Two systems have been trialled.
The first is a health management information system (HMIS) based on DHIS2, a web-based opensource information system. The solution strategy is based on data governance and management
using i* (iStar) methodology and alignment with the DHIS2 ecosystem for sustainability and
evolution. All levels of health data use the same software, online or offline. The second health
information system (HIS) uses DHIS2 set up as a cloud-based centralised repository and a custom
application, Praxis, a Chrome application for project-level users that operates on and offline and
enables uploading of data and downloading of analyses of programme indicators. We report on the
performance of both systems.
Methods
HMIS was deployed in 60 projects in 16 countries from August to April 2016; 350 users were trained.
Agile/SCRUM methodologies were applied in all tracks of the project, allowing streamlined
implementation (level of resources used, time to deliver). A mobile implementation team of two
people spent an average of 3 weeks per country, going to each project to set up the system (data
flow and data to be collected) and train users. Praxis was deployed in 17 projects, including eight
hospitals, in nine countries between September and April 2016. A mobile implementation team of
four people spent an average of 14 days per project to set up data flows, paper registers, and train
users. Effectiveness of HIS/Praxis was assessed through the System Usability Scale (SUS; measuring
user ability in project setup, data entry and approval, and import/export of the field application),
and timely data submission over 2 weeks (weekly data submission was used as a proxy to determine
uptake of the system and access to timely data), and response to HIS support queries.
Ethics
This description/evaluation of an innovation project involved human participants or their data, and
has had ethics oversight from the medical director or delegated representative according to the MSF
Ethics Framework for Innovation.
Results
237 evaluations of HMIS training were conducted; 83% (197) of users rated it as very useful or
useful. 74% (175) of users were satisfied with HMIS implementation. Implementation was planned
for 15 countries from July to March; the result was 16 countries from August to April. The human
resources planned and devoted to deployment were four people with a health profile and four
software engineers. For Praxis, SUS was 75 (68 is a good score). In the two assessed weeks (ISO
weeks 14 and 15, 2016), 93% (14/15) and 88% (14/16) of projects submitted weekly data on time.
During the first 6 months of deployment, of 35 support requests: nine tier-1 functional requests
were resolved within 1 day; six tier-1 technical within 3 days; and 17 tier-3 technical within 10 days.
Conclusion
HMIS facilitates access to health information at all levels, maximizing the decision-making process.
Field teams are empowered through effective health data and information management. The next
step is integrating complementary functionalities for specific services or contexts. Praxis deployment
has shown that project, coordination, and headquarters users now have timely access to project
data. We expect in the coming months that this availability will accelerate the decision-making
process. The deployment has also highlighted challenges in infrastructure, information management,
and expectations of technology deployment. Praxis will be rolled out further alongside monitoring
and evaluation support teams to help projects understand and use project data in decision-making.
Conflicts of interest
None declared.